High Performance Data Visualization for the Web

on YouTube ↗29m

" summary

Built a tiny FastAPI site and client that wires Perspective (C++/Rust engine compiled to WebAssembly) into a browser web component and a Jupyter widget, demonstrating a datagrid, heatmaps and maps with cross-filtering and computed columns and live updates. Data flows over websockets using Apache Arrow diffs; demo streams tick at 1s then 0.1s and is pushed to 0.01s successfully. Perspective is Apache‑2, handles several gigabytes for flat views but heavy pivots/aggregations slow down. FastAPI + websockets with Arrow diffs and Perspective (C++/Rust, WebAssembly).

AI-written summary · about

" chapters (7)

1. 0:00Introduction to Perspective
2. 1:03Data challenges on the web
3. 5:53Perspective library overview
4. 9:30Interactive demos and features
5. 14:12Python widget implementation
6. 18:15Web application integration
7. 21:47Q&A and closing remarks

" related talks

• 28mHigh Performance Data Visualization for the WebTim Paine · PyData Paris 2024
• 20mA framework for building data driven interactive front endsRitvvij Parrikh · Meta Refresh 2015
• 49mStreaming VisualisationGuido Schmutz · Voxxed Days Zurich 2019
• 46mExplorations in Interactive Visual AnalyticsDoug Talbott · GOTO Copenhagen 2014
• 27mScalable Data VisualizationZach Nation · React.js Conf 2015
• 36mMaking interactive visualizations easy, inside and outside of notebooksJulia Signell · PyData DC 2018
• 43mThe Future of Data Visualization on the WebAlan Mendelevich · NDC Oslo 2018
• 27mStream, Store, Visualize: Fast-Track Real-Time Analytics the OSS WayKamesh Sampath · Great International Developer Summit 2025