Buffer-Level NETS Processing #25

NEXT MOVES:
PROCESS GIS MEASURES FOR EXAM 6
disaster nets
food stores nets (include alcohol for mariana - steve 2/27)
social nets
calculate distances at 5mi threshold, then group non spatially

Outline

use jupyter notebook in arcgis

1. use generate near table to join mesa locations to duns locations spatial file at a five mile threshold
2. join table back to duns location on FIDs to get AddressID
   Columns: LOCID_DREXEL, AddressID
3. join dunsmove table to get DunsYear, year (outer join? or dunsmove on left, left join?)
   Columns: LOCID_DREXEL, AddressID, DunsYear, Year
4. join classifiedlong table to get BaseGroup
   Columns: LOCID_DREXEL, AddressID, DunsYear, Year, BaseGroup
5. join xwalk to get high level categories
   Columns: LOCID_DREXEL, AddressID, DunsYear, Year, BaseGroup, HighLevel

what format do we need output to be in? can it be long in this format? how will analysts be using the output? do they want it to be separated by year?

@kam642 @sjmelly I've tested processing on all MESA participant locations with UHCMatchCodeRank <=6, using all five buffer sizes, and using categories in the domains 'Food' and 'Disaster/Construction', plus 'ACP'. I have a script that will take various inputs such as categories, domains, UHCMatchCodeRank (for both BEDDN and MESA locs), and years and will produce datasets with each category as a column, with each row as a buffer-year, with category counts as values. The script also saves the MESA/BEDDN distances from the generate near table in chunks, for anyone interested in looking at actual distances instead of just counts in buffers.

Code

https://github.com/Drexel-UHC/MESA_Neigh_Aging/tree/main/nets_linkage
mesa_to_beddn_generatenear.ipynb loops through MESA locs in chunks of 1000 and calculates distances to each BEDDN loc. this has been run once and does not need to be run again.
mesa_to_beddn_linkcats.ipynb loops through the generate near tables and mesabuff_beddn_yyyy.csv(s).

Output

Y:\MESA_Neigh_Age\Data\NETS\mesa_linkage_test
near_with_ids this folder contains the NEAR_DIST output, with the unique ids joined back in. The MESA locs were processed in chunks of 1000, so there are 17 files.
measures This folder contains files that have counts of categories per buffer-year. Files are divided up into years.
mesabuff_beddn_metadata contains metadata from the mesa_to_beddn_linkcats process
mesabuff_beddn_stats contains descriptives from the mesa_to_beddn_linkcats process

Next Steps: process all categories for the years processed above (2000, 2005, 2010). If this works, we can continue to process yearly for all categories.

Buffer level MESA/BEDDN linkage is done. Y:\MESA_Neigh_Age\Data\Measures\MESABuffer\beddn_mesa_links contains files for every year with counts for all 255 categories at all five buffer levels, long by buffer level. Y:\MESA_Neigh_Age\Data\Measures\MESABuffer\beddn_mesa_links_stats has metadata files and an excel doc with sheets containing descriptive stats for each year. @kam642 @sjmelly what checks would you like to see done on my end?