📊 Shiny for R on Databricks Apps

A sample Shiny app written in R that runs natively on Databricks Apps — deployed via DABs (Databricks Asset Bundles).

🤔 What's the approach?

Databricks Apps instances are shipped with Python and uv, but unfortunately doesn't have a preinstalled R runtime.

This project shows how to use a lightweight Python launcher that bootstraps an R environment at startup using micromamba, then hands off execution to a Shiny app.

🏗️ How it works

Databricks App Instance
=========================

uv run scripts/launcher.py
  |-- 1. Downloads micromamba
  |-- 2. Creates conda env from environment.yml
  |-- 3. Installs R + packages
  '-- 4. Starts Shiny app via Rscript

app/app.R  -->  serves on $DATABRICKS_APP_PORT

The Shiny app queries Databricks SQL using the forwarded user token (HTTP_X_FORWARDED_ACCESS_TOKEN), so each user sees data scoped to their own permissions.

📁 Project structure

├── app/
│   └── app.R                # Shiny app (UI + server)
├── scripts/
│   └── launcher.py          # Python bootstrap script
├── environment.yml          # Conda env: R + packages
└── databricks.yml           # DABs bundle definition

✅ Prerequisites

• Databricks CLI installed and authenticated
• A Databricks SQL Warehouse ID
• Access to a Databricks workspace

🚀 Deploy

1. Validate the bundle

databricks bundle validate

2. Deploy to your workspace

databricks bundle deploy -var="sql_warehouse_id=<YOUR_WAREHOUSE_ID>"

3. Launch the app

databricks bundle run shiny-app-r

That's it — the app will be live on your workspace's Apps URL. 🎉

🔧 Configuration

The bundle is defined in databricks.yml. Key settings:

Setting	Value
App name	shiny-app-r
Start command	uv run scripts/launcher.py
User API scopes	sql
Variable	sql_warehouse_id — passed as SQL_WAREHOUSE_ID env var

R dependencies live in environment.yml and are installed at container startup via micromamba.

🔑 Authentication

Databricks Apps handle auth automatically. The app reads the user's token from the X-Forwarded-Access-Token header injected by the Apps proxy — no secrets management needed.

💡 Adapting this for your own Shiny app

1. Replace app/app.R with your Shiny application
2. Add any new R packages to environment.yml
3. Update the SQL query or data source as needed
4. Re-deploy with databricks bundle deploy

📚 References

• Databricks Apps
• micromamba
• Shiny
• R
• uv