Usage Overview¶
Spark Plan Viz provides three ways to work with execution plans:
| Mode | Best for | Function |
|---|---|---|
| Notebook Mode | Interactive exploration in Jupyter | visualize_plan(df, notebook=True) |
| File Mode | Sharing HTML with teammates | visualize_plan(df, notebook=False) |
| Programmatic API | CI pipelines, automated checks | analyze_plan(df) |
All modes parse the same underlying JVM execution plan via Py4J and run the same 15-rule optimization analysis.