Streamlit Fundamentals and Data App Labs

Practice Streamlit basics and build a small interactive task board in three short Python lab scripts (no notebooks), plus an end-to-end ML → API → Streamlit exercise.

Environment setup

1. Python 3.11 recommended (optional): pyenv local 3.11.3
2. Create/activate a virtual environment:
   • macOS/Linux: python -m venv .venv && source .venv/bin/activate
   • Windows (PowerShell): python -m venv .venv && .venv\\Scripts\\Activate.ps1
3. Install deps: pip install --upgrade pip && pip install -r requirements.txt

Contents (suggested order)

1. 01_streamlit_fundamentals.py — First steps with Streamlit, page layout, text, inputs, and reruns.
2. 02_streamlit_widgets_state.py — Forms, session state, callbacks, and handling user input.
3. 03_streamlit_data_app.py — Building a simple data app with tables, filtering, and tidy UI patterns.
4. main.py — Task-board app that combines the lab patterns (add/filter/complete/delete tasks).
5. ml_api_demo/ — Self-contained ML → API → Streamlit exercise (notebook + redundant script for headless retraining):
   • 04_train_and_save_model.ipynb — Train and save a simple iris classifier (produces ml_api_demo/artifacts/iris_model.joblib).
   • train_model.py — Script alternative to produce the same artifact without notebooks.
   • model_api.py — FastAPI service that loads the saved model and exposes /predict.
   • streamlit_api_client.py — Streamlit UI to call the API and view predictions.
   • More detail: see ml_api_demo/README.md.

Run the Streamlit labs (UI patterns)

• Fundamentals: streamlit run 01_streamlit_fundamentals.py
• Widgets + State: streamlit run 02_streamlit_widgets_state.py
• Data App: streamlit run 03_streamlit_data_app.py
• Task board demo: streamlit run main.py

End-to-end ML → API → Streamlit (in ml_api_demo/)

1. Train and save the model
   • Notebook: run ml_api_demo/04_train_and_save_model.ipynb
   • Script: python ml_api_demo/train_model.py
   • Output: ml_api_demo/artifacts/iris_model.joblib
2. Start the API
   • uvicorn ml_api_demo.model_api:app --reload --port 8001
   • Test:

     curl -X POST http://127.0.0.1:8001/predict \
       -H "Content-Type: application/json" \
       -d '{"sepal_length":5.1,"sepal_width":3.5,"petal_length":1.4,"petal_width":0.2}'

3. Query from Streamlit
   • streamlit run ml_api_demo/streamlit_api_client.py
   • Point the client at http://127.0.0.1:8001 and send predictions.