Update to 2024 versions of black and scikit-learn (#504)

* switch from black 23 to black 24

* fix mfkpls

setup.py

@@ -7,6 +7,7 @@
         
 This package is distributed under New BSD license.
 """
+
 from setuptools import setup, Extension
 import sys
 import numpy as np

smt/applications/mfkpls.py

@@ -12,7 +12,7 @@
 import numpy as np
 
 from sklearn.cross_decomposition import PLSRegression as pls
-from sklearn.metrics.pairwise import manhattan_distances
+from sklearn.metrics.pairwise import check_pairwise_arrays
 
 from smt.applications import MFK
 from smt.utils.kriging import componentwise_distance_PLS
@@ -42,7 +42,10 @@ def _differences(self, X, Y):
         Overrides differences function for MFK
         Compute the manhattan_distances
         """
-        return manhattan_distances(X, Y, sum_over_features=False)
+        X, Y = check_pairwise_arrays(X, Y)
+        D = X[:, np.newaxis, :] - Y[np.newaxis, :, :]
+        D = np.abs(D, D)
+        return D.reshape((-1, X.shape[1]))
 
     def _componentwise_distance(self, dx, opt=0):
         d = componentwise_distance_PLS(

smt/applications/mixed_integer.py

@@ -215,9 +215,9 @@ def __init__(
             )
             and self._surrogate.options["categorical_kernel"] is None
         ):
-            self._surrogate.options[
-                "categorical_kernel"
-            ] = MixIntKernelType.HOMO_HSPHERE
+            self._surrogate.options["categorical_kernel"] = (
+                MixIntKernelType.HOMO_HSPHERE
+            )
             warnings.warn(
                 "Using MixedIntegerSurrogateModel integer model with Continuous Relaxation is not supported. Switched to homoscedastic hypersphere kernel instead."
             )

smt/applications/moe.py

@@ -5,6 +5,7 @@
 
 Mixture of Experts
 """
+
 # TODO : support for best number of clusters
 # TODO : implement verbosity 'print_global'
 # TODO : documentation

smt/examples/multi_modal/run_genn_demo.py

@@ -9,6 +9,7 @@
 
 This package is distributed under New BSD license.
 """
+
 from smt.surrogate_models.genn import GENN, load_smt_data
 
 import numpy as np

smt/problems/branin.py

@@ -5,6 +5,7 @@
 
 Branin function.
 """
+
 import numpy as np
 
 from smt.problems.problem import Problem

smt/problems/cantilever_beam.py

@@ -7,6 +7,7 @@
 Liu, H., Xu, S., & Wang, X. Sampling strategies and metamodeling techniques for engineering design: comparison and application. In ASME Turbo Expo 2016: Turbomachinery Technical Conference and Exposition. American Society of Mechanical Engineers. June, 2016.
 Cheng, G. H., Younis, A., Hajikolaei, K. H., and Wang, G. G. Trust Region Based Mode Pursuing Sampling Method for Global Optimization of High Dimensional Design Problems. Journal of Mechanical Design, 137(2). 2015.
 """
+
 import numpy as np
 
 from smt.problems.problem import Problem
@@ -75,16 +76,10 @@ def _evaluate(self, x, kx):
                 b = x[:, 3 * kelem + 0]
                 h = x[:, 3 * kelem + 1]
                 y[:, 0] += (
-                    -12.0
-                    / b**2
-                    / h**3
-                    * np.sum(x[:, 2 + 3 * kelem :: 3], axis=1) ** 3
+                    -12.0 / b**2 / h**3 * np.sum(x[:, 2 + 3 * kelem :: 3], axis=1) ** 3
                 )
                 y[:, 0] -= (
-                    -12.0
-                    / b**2
-                    / h**3
-                    * np.sum(x[:, 5 + 3 * kelem :: 3], axis=1) ** 3
+                    -12.0 / b**2 / h**3 * np.sum(x[:, 5 + 3 * kelem :: 3], axis=1) ** 3
                 )
 
             elif kx % 3 == 1:
@@ -101,10 +96,7 @@ def _evaluate(self, x, kx):
                     b = x[:, 3 * ielem + 0]
                     h = x[:, 3 * ielem + 1]
                     y[:, 0] += (
-                        36.0
-                        / b
-                        / h**3
-                        * np.sum(x[:, 2 + 3 * ielem :: 3], axis=1) ** 2
+                        36.0 / b / h**3 * np.sum(x[:, 2 + 3 * ielem :: 3], axis=1) ** 2
                     )
                     if kelem > ielem:
                         y[:, 0] -= (

smt/problems/hierarchical_goldstein.py

@@ -5,6 +5,7 @@
 Cantilever beam problem from:
 P. Saves, Y. Diouane, N. Bartoli, T. Lefebvre, and J. Morlier. A mixed-categorical correlation kernel for gaussian process, 2022
 """
+
 import numpy as np
 
 from smt.problems.problem import Problem

smt/problems/lp_norm.py

@@ -5,6 +5,7 @@
 
 Norm function.
 """
+
 import numpy as np
 
 from smt.problems.problem import Problem

smt/problems/mixed_cantilever_beam.py

@@ -5,6 +5,7 @@
 Cantilever beam problem from:
 P. Saves, Y. Diouane, N. Bartoli, T. Lefebvre, and J. Morlier. A mixed-categorical correlation kernel for gaussian process, 2022
 """
+
 import numpy as np
 
 from smt.problems.problem import Problem

smt/problems/ndim_cantilever_beam.py

@@ -5,6 +5,7 @@
 
 N-dimensional cantilever beam problem.
 """
+
 import numpy as np
 
 from smt.utils.options_dictionary import OptionsDictionary

smt/problems/ndim_robot_arm.py

@@ -5,6 +5,7 @@
 
 N-dimensional robot arm problem.
 """
+
 import numpy as np
 
 from smt.utils.options_dictionary import OptionsDictionary

smt/problems/ndim_rosenbrock.py

@@ -5,6 +5,7 @@
 
 N-dimensional Rosenbrock problem.
 """
+
 import numpy as np
 
 from smt.utils.options_dictionary import OptionsDictionary

smt/problems/ndim_step_function.py

@@ -5,6 +5,7 @@
 
 N-dimensional step function problem.
 """
+
 import numpy as np
 
 from smt.utils.options_dictionary import OptionsDictionary

smt/problems/neural_network.py

@@ -6,6 +6,7 @@
 C. Audet, E. Hall e-Hannan, and S. Le Digabel. A general mathematical framework for constrained mixed-variable blackbox optimization problems with meta and categorical variables. Operations Research Forum,499
 4:137, 2023.
     """
+
 import numpy as np
 
 from smt.problems.problem import Problem

smt/problems/problem.py

@@ -5,6 +5,7 @@
 
 Base class for benchmarking/test problems.
 """
+
 from typing import Optional
 import numpy as np
 

smt/problems/reduced_problem.py

@@ -5,6 +5,7 @@
 
 Reduced problem class - selects a subset of input variables.
 """
+
 import numpy as np
 
 from smt.utils.options_dictionary import OptionsDictionary

smt/problems/robot_arm.py

@@ -7,6 +7,7 @@
 Liu, H., Xu, S., & Wang, X. Sampling strategies and metamodeling techniques for engineering design: comparison and application. In ASME Turbo Expo 2016: Turbomachinery Technical Conference and Exposition. American Society of Mechanical Engineers. June, 2016.
 An, J., and Owen, A. Quasi-Regression. Journal of complexity, 17(4), pp. 588-607, 2001.
 """
+
 import numpy as np
 
 from smt.problems.problem import Problem

smt/problems/rosenbrock.py

@@ -5,6 +5,7 @@
 
 Multi-dimensional Rosenbrock function.
 """
+
 import numpy as np
 
 from smt.problems.problem import Problem

smt/problems/sphere.py

@@ -6,6 +6,7 @@
 
 Sphere function.
 """
+
 import numpy as np
 
 from smt.problems.problem import Problem

smt/problems/tensor_product.py

@@ -5,6 +5,7 @@
 
 Tensor-product of cos, exp, or tanh.
 """
+
 import numpy as np
 
 from smt.problems.problem import Problem

smt/problems/torsion_vibration.py

@@ -8,6 +8,7 @@
 Liu, H., Xu, S., & Wang, X. Sampling strategies and metamodeling techniques for engineering design: comparison and application. In ASME Turbo Expo 2016: Turbomachinery Technical Conference and Exposition. American Society of Mechanical Engineers. June, 2016.
 Wang, L., Beeson, D., Wiggs, G., and Rayasam, M. A Comparison of Metamodeling Methods Using Practical Industry Requirements. In Proceedings of the 47th AIAA/ASME/ASCE/AHS/ASC structures, structural dynamics, and materials conference, Newport, RI, pp. AIAA 2006-1811.
 """
+
 import numpy as np
 from scipy.misc import derivative
 

smt/problems/water_flow.py

@@ -8,6 +8,7 @@
 Liu, H., Xu, S., & Wang, X. Sampling strategies and metamodeling techniques for engineering design: comparison and application. In ASME Turbo Expo 2016: Turbomachinery Technical Conference and Exposition. American Society of Mechanical Engineers. June, 2016.
 Morris, M. D., Mitchell, T. J., and Ylvisaker, D. Bayesian Design and Analysis of Computer Experiments: Use of Derivatives in Surface Prediction. Technometrics, 35(3), pp. 243-255. 1993.
 """
+
 import numpy as np
 from scipy.misc import derivative
 

smt/problems/water_flow_lfidelity.py

@@ -5,6 +5,7 @@
 Water flow problem from:
 Xiong, S., Qian, P. Z., & Wu, C. J. (2013). Sequential design and analysis of high-accuracy and low-accuracy computer codes. Technometrics, 55(1), 37-46.
 """
+
 import numpy as np
 from scipy.misc import derivative
 

smt/problems/welded_beam.py

@@ -8,6 +8,7 @@
 Liu, H., Xu, S., & Wang, X. Sampling strategies and metamodeling techniques for engineering design: comparison and application. In ASME Turbo Expo 2016: Turbomachinery Technical Conference and Exposition. American Society of Mechanical Engineers. June, 2016.
 Deb, K. An Efficient Constraint Handling Method for Genetic Algorithms. Computer methods in applied mechanics and engineering, 186(2), pp. 311-338. 2000.
 """
+
 import numpy as np
 from scipy.misc import derivative
 from smt.problems.problem import Problem

smt/problems/wing_weight.py

@@ -9,6 +9,7 @@
 Forrester, A., Sobester, A., and Keane, A., 2008,
 Engineering Design Via Surrogate Modelling: A Practical Guide, John Wiley & Sons, United Kingdom.
 """
+
 import numpy as np
 from scipy.misc import derivative
 

smt/sampling_methods/full_factorial.py

@@ -5,6 +5,7 @@
 
 Full-factorial sampling.
 """
+
 import numpy as np
 
 from smt.sampling_methods.sampling_method import ScaledSamplingMethod

smt/sampling_methods/lhs.py

@@ -5,6 +5,7 @@
 
 LHS sampling; uses the pyDOE3 package.
 """
+
 from pyDOE3 import lhs
 from scipy.spatial.distance import pdist, cdist
 import numpy as np
@@ -290,12 +291,8 @@ def _PhiP_exchange(self, X, k, PhiP_, p, fixed_index):
 
         dist1 = cdist([X[i1, :]], X_)
         dist2 = cdist([X[i2, :]], X_)
-        d1 = np.sqrt(
-            dist1**2 + (X[i2, k] - X_[:, k]) ** 2 - (X[i1, k] - X_[:, k]) ** 2
-        )
-        d2 = np.sqrt(
-            dist2**2 - (X[i2, k] - X_[:, k]) ** 2 + (X[i1, k] - X_[:, k]) ** 2
-        )
+        d1 = np.sqrt(dist1**2 + (X[i2, k] - X_[:, k]) ** 2 - (X[i1, k] - X_[:, k]) ** 2)
+        d2 = np.sqrt(dist2**2 - (X[i2, k] - X_[:, k]) ** 2 + (X[i1, k] - X_[:, k]) ** 2)
 
         res = (
             PhiP_**p + (d1 ** (-p) - dist1 ** (-p) + d2 ** (-p) - dist2 ** (-p)).sum()

smt/sampling_methods/pydoe.py

@@ -5,6 +5,7 @@
 
 pyDOE3 sampling methods
 """
+
 from pyDOE3 import doe_box_behnken
 from pyDOE3 import doe_gsd
 from pyDOE3 import doe_factorial

smt/sampling_methods/random.py

@@ -5,6 +5,7 @@
 
 Random sampling.
 """
+
 import numpy as np
 
 from smt.sampling_methods.sampling_method import ScaledSamplingMethod

smt/sampling_methods/sampling_method.py

@@ -5,6 +5,7 @@
 
 Base class for sampling algorithms.
 """
+
 from abc import ABCMeta, abstractmethod
 import numpy as np
 import warnings

smt/surrogate_models/idw.py

@@ -4,6 +4,7 @@
 
 This package is distributed under New BSD license.
 """
+
 import numpy as np
 from scipy.sparse import csc_matrix
 from smt.surrogate_models.surrogate_model import SurrogateModel
@@ -13,7 +14,6 @@
 
 
 class IDW(SurrogateModel):
-
     """
     Inverse distance weighting interpolant
     This model uses the inverse distance between the unknown and training

smt/surrogate_models/krg_based.py

@@ -3,6 +3,7 @@
 Some functions are copied from gaussian_process submodule (Scikit-learn 0.14)
 This package is distributed under New BSD license.
 """
+
 import numpy as np
 from enum import Enum
 from scipy import linalg, optimize

smt/surrogate_models/ls.py

@@ -7,6 +7,7 @@
 TO DO:
 - define outputs['sol'] = self.sol
 """
+
 import numpy as np
 
 from sklearn import linear_model
@@ -15,7 +16,6 @@
 
 
 class LS(SurrogateModel):
-
     """
     Least square model.
     This model uses the linear_model.LinearRegression class from scikit-learn.

smt/surrogate_models/qp.py

@@ -7,6 +7,7 @@
 TO DO:
 - define outputs['sol'] = self.sol
 """
+
 import numpy as np
 import scipy
 from smt.surrogate_models.surrogate_model import SurrogateModel
@@ -15,7 +16,6 @@
 
 
 class QP(SurrogateModel):
-
     """
     Square polynomial approach
     """

smt/surrogate_models/rbf.py

@@ -3,6 +3,7 @@
 
 This package is distributed under New BSD license.
 """
+
 import numpy as np
 from scipy.sparse import csc_matrix
 from smt.surrogate_models.surrogate_model import SurrogateModel
@@ -14,7 +15,6 @@
 
 
 class RBF(SurrogateModel):
-
     """
     Radial basis function interpolant with global polynomial trend.
     """

smt/surrogate_models/rmtb.py

@@ -3,6 +3,7 @@
 
 This package is distributed under New BSD license.
 """
+
 import numpy as np
 import scipy.sparse
 from numbers import Integral

smt/surrogate_models/rmtc.py

@@ -3,6 +3,7 @@
 
 This package is distributed under New BSD license.
 """
+
 import numpy as np
 import scipy.sparse
 from numbers import Integral