Understanding reactive execution in Shiny
summary
Shiny for Python claims to infer component dependencies and minimally rerender by building a computation graph via runtime tracing rather than manual callbacks; background tabs do not compute (so 100 tabs cost nothing until opened), reactive calculations use reactive calc to cache shared intermediate results, and reactive event/effect handles explicit side effects. The framework runs entirely in Python (no JS), supports dynamic UIs, demonstrates cached sampling+filtering that shares one dataset between plots, and scales from quick demos to production apps.
AI-written summary · about
chapters (7)
- 0:00Shiny introduction
- 3:49Frameworks scale with growth
- 8:01Shiny reactivity rendering
- 14:17Reactive graphs complexity
- 17:46Reactive calculations efficiency
- 21:16Side effects beyond reactivity
- 24:46Summary
related talks
35mBuild Simple and Scalable Apps with Shiny
44mShiny: Data-centric web applications in Python
36mReactive Notebooks for Python
20mRunning Shiny without a server- 20mRunning Shiny without a server
20mSupercharge Your Shiny (for Python) App: Unleashing Interactive Jupyter Widgets
20mWeb applications with Shiny and React and AI
01h 44mShinylive - Running R and Python Dashboards without a Server