Shiny: Data-centric web applications in Python
summary
Shiny for Python is a ground-up reimplementation (not an R wrapper) that uses Transparent Reactive Programming to build data apps with Seaborn, pandas/polars, Plotly and media‑pipe hand tracking. Demos show ChatGPT streaming via async generators, a working 20‑question self‑play demo, and a camera‑based plotly view that becomes smooth by smoothing over five samples. ShinyLive runs wholly in the browser via Pyodide, reactive caching holds one in‑memory copy per session, and the draggable UI editor is upcoming.
AI-written summary · about
chapters (20)
- 0:00Shiny for Python intro
- 1:48Python vs R Shiny
- 3:36Basic Shiny app structure
- 5:44Complex Shiny UI
- 7:54Streamlit comparison
- 12:09Reactive programming
- 13:54LLM integration demo
- 17:50ChatGPT chat demo
- 20:303D hand tracking demo
- 24:10ShinyLive: WebAssembly
- 27:17Quarto integration
- 28:02UI editor preview
- 29:26Resources & closing
- 30:07Q&A: Large datasets
- 33:17Q&A: R to Python porting
- 34:25Q&A: Dash vs Shiny
- 37:06Q&A: Bokeh & Flask
- 38:14Q&A: Multi-user sessions
- 40:24Q&A: Data security
- 41:27Q&A: Deployment & CI/CD
related talks
44mShiny for Python: Interactive apps and dashboards made easy-ish
35mBuild Simple and Scalable Apps with Shiny
31mTips to Level-Up Your Shiny for Python Applications
25mUnderstanding reactive execution in Shiny
20mRunning Shiny without a server- 20mRunning Shiny without a server
01h 44mShinylive - Running R and Python Dashboards without a Server
27mShiny: A GUI for the Tidyverse