Colorado Public Health — R Shiny (Socrata SODA v2)

Interactive Shiny app that queries Colorado’s open-data Socrata SODA v2 endpoint (/resource/<VIEW_ID>.json using $query) and shows descriptive stats by date/group.

• Default dataset: a498-a63g (you can swap the view id anytime).
  Note: this demo view doesn’t have a true date column; the time-series will be blank until you use a dataset with dates.

---

Requirements

• R (≥ 4.2)
• Internet access to https://data.colorado.gov
• (Recommended) a Socrata App Token

---

1) Get the code

# first time
git clone https://github.com/ileleiwi/co-shiny-socrata.git
cd co-shiny-socrata

# later, to update
git pull

---

2) Install R packages

R -q -e "source('install.R')"

---

3) Set your Socrata App Token

Create a token at https://data.colorado.gov/profile/edit/developer_settings, then set it as an environment variable:

macOS / Linux (bash/zsh)

export SOCRATA_APP_TOKEN="your_token_here"

Windows PowerShell

$env:SOCRATA_APP_TOKEN = "your_token_here"

R (temporary for this session)

Sys.setenv(SOCRATA_APP_TOKEN = "your_token_here")

The app reads the token from config.yml → app_token_env (defaults to SOCRATA_APP_TOKEN).
The sidebar will show “Token detected ✓” when it’s available.

---

4) Run the app locally

R -q -e "shiny::runApp(port=8080, host='127.0.0.1', launch.browser=TRUE)"

If your browser doesn’t open automatically, visit: http://localhost:8080

---

5) Use the app

1. Click Health check → should report row/column counts.
2. In the SoQL box, start with:

   SELECT * LIMIT 100

3. Click Fetch.
4. In Pickers:
   • Choose a Group column (e.g., race_ethnicity).
   • Choose a Numeric column (e.g., median_household_income).
   • (Optional) Choose a Date column + range if your dataset has dates.
5. Explore:
   • Time series (sum per date; requires a date column).
   • Grouped stats bar chart (Sum / Mean / Median / Count).
   • Grouped summary table and Raw data (both downloadable as CSV).

---

6) Switch datasets

Two ways:

• Quick (UI only): Replace Socrata View ID in the sidebar (e.g., abcd-1234) and adjust your SoQL.
• Config (persisted): Edit config.yml:

  socrata_domain: data.colorado.gov
  view_id: a498-a63g
  app_token_env: SOCRATA_APP_TOKEN

Tip: Use SoQL for server-side aggregation, e.g.:

SELECT date_trunc_ymd(date_col) AS day, group_col, SUM(value_col) AS n
GROUP BY day, group_col
ORDER BY day
LIMIT 50000

---

Troubleshooting

• Token shows “missing” → ensure the env var is set in the same shell you launch R from.
• HTTP 403 → token not sent or invalid.
• HTTP 400 → SoQL syntax error. Try SELECT * LIMIT 10.
• Port busy → change the port, e.g. port=8888.
• Nothing loads → open http://localhost:8080 manually.

---

Project structure (key files)

• app.R — Shiny UI/server (theme, grouped stats plot, downloads)
• R/socrata_v2.R — robust SODA v2 client (raw→UTF-8, SoQL paging)
• install.R — installs CRAN deps
• config.yml — domain, default view_id, and token env var

---

License

MIT