(R workspace) Load Libraries:

• caret
• data.table
• dplyr
• sets
• scales
• tidyr
• stringr

(Python workspace) Load packages:

• json
• pandas

• After downloading Data folder from specified OneDrive:

.
├── AHRQ_pipeline
│   ├── compress_pharma.py
│   ├── merge_AHRQ.Rmd
│   ├── gp_zipdata.json
│   └── fsq_zipdata.json
└── Data: download from OneDrive
    ├── AHRQ
    │   ├── COUNTY
    │   ├── TRACT
    │   └── ZIP
    └── GA_Pharmacy_Data_gp_fsq

1. merge_AHRQ.Rmd: reads all available years of AHRQ SDOH data at ZIP code level. Generates summary plots, and includes template to merge SODH with patient data by ZIP code, year.
2. compress_pharma.py: reads Georgia pharmacy locations, and generates summary statistics by writing a json file with zip codes (key) and the number of pharmacies in that zip code (value).

Merge the preprocessed AHRQ data with patient data using crosswalk variables:

• STATEFIPS
• ZIPCODE
• YEAR Merge the Georgia pharmacy location data with patient data using crosswalk variables:
• zipcode
• can optionally select either 'gp' or 'fsq' as the data source for Georgia pharmacy location data (see gp_zipdata.json, fsq_zipdata.json), or use both sources combined (

• AHRQ SDOHD: https://www.ahrq.gov/sdoh/data-analytics/sdoh-data.html#download (If you have access, we recommend using the OneDrive ZIP code level data because we have already isolated relevant sheets from the csv files at the above link)