fastapi-penguins

demo-fastapi-penguins

• Code: https://github.com/SamEdwardes/demo-fastapi-penguins
• Deployment: https://colorado.rstudio.com/rsc/demo-fastapi-penguins/docs

Usage

Create a virtual environment.

uv sync

Then run the app.

uv run uvicorn app:app --reload

You could also call the API using curl:

CONNECT_SERVER='https://pub.current.posit.team'

curl -L -X 'GET' \
  "${CONNECT_SERVER}/public/fastapi-penguins/penguins?sample_size=1" \
  -H 'accept: application/json'

If the API is private you will need to authenticate using your API key.

CONNECT_SERVER='https://pub.current.posit.team'

curl -L -X 'GET' \
  "${CONNECT_SERVER}/public/fastapi-penguins/penguins?sample_size=1" \
  -H 'accept: application/json' \
  -H "Authorization: Key ${CONNECT_API_KEY}"

Cleaner syntx with httpie:

uvx --from httpie https pub.current.posit.team/public/fastapi-penguins/penguins \
  "Authorization:Key ${CONNECT_API_KEY}" \
  sample_size==2

Deployment

Git-backed

The app is automatically deployed to RStudio connect using git backed deployment. Make any changes to the code, then run the following:

uv run export -o requirements.txt --no-hashes

rsconnect write-manifest fastapi \
  --overwrite \
  --python .venv/bin/python \
  --entrypoint app:app \
  .

⚠️ Remember to update the app/requirements.txt file if you add any new packages.

Programmatic

You can deploy the app using the rsconnect cli:

uv run export -o requirements.txt --no-hashes

rsconnect deploy fastapi \
  --python .venv/bin/python \
  --entrypoint app:app \
  --new \
  .