issues with integrating new microdata with urbansim

[fscottfoti]

Critical modeling issues

• x and y on the parcels
• CH the spatial joins take a long time in merge_cities - move these to another script
• CH add dummy parcels so dummy buildings have something to join to
• CH deal with building types urbansim doesn't know about
• CH move geometry out of buildings so it's faster to read in buildings
• go back to ints for parcel ids and building ids, not worth fixing urbansim to deal with this
• write a script to package up the outputs of cutting_board as parcels.csv, buildings.csv, households.csv, and jobs.csv
• need to re-estimate elcms (for some reason takes a very long time to process
• there was an issue with add_age_categories in the summary - drop for now
• "static_parcels", manual edits, firebase nodevs, and development projects model need to be revived (geom_id dependency)
• MR development projects should probably just have x/y which links to parcel. many of the x's and y's are reversed in the file.
• no longer have nodevs in parcels_geography.csv? also make the sdem column on parcels in variables.py work again
• dropping gq households in preprocessing and in households wrapper
• jobs relocation commented out for empsix issues
• the describe of price always sets residential to 5.178

Finish up

• summaries.py hardcodes sectors in two places (search for AGREMPN) - these were required by the old travel model - I assume we need the new categories now
• add maz summaries, make sure we generate what the travel model needs

Cleanup

• move cutting_board into basis
• generate residential units and deed restricted units in cutting_board/basis instead of bayarea_urbansim
• add script to cutting_board which runs the final steps (merge_cities, assign_households, assign_jobs)
• get rid of parcels_geography - all geographies now on parcels
• drop old data files