Seeing the needle AND the haystack: single-datapoint selection for billion-poin...

on YouTube ↗30m

" summary

Claims server‑side rendering handles billions by rasterizing aggregates and shipping counts to the browser, using Datashader, with client colormapping to preserve histogram‑equalized dynamic range while avoiding overplotting. The demo combines hvPlot + HoloViews + Bokeh + Panel, shows 2.4M gull samples and ~200M AIS ship pings and references the ~330M US census, supports hover/inspect down to a single datapoint, produces deployable Bokeh dashboards, and recommends Parquet/Dask and spatial indexing for performance.

AI-written summary · about

" chapters (10)

1. 0:00Intro
2. 2:28Data Shader
3. 5:02Data Shader Demo
4. 6:09Interactive Plotting
5. 7:38H3 Plot
6. 11:09H3 Color Bar
7. 12:20Inspect
8. 15:27Dashboard
9. 16:44Other features
10. 19:42Ship traffic

" related talks

• 46mExplorations in Interactive Visual AnalyticsDoug Talbott · GOTO Copenhagen 2014
• 02h 14mData Applications with BokehBryan Van de Ven · PyData San Francisco 2016
• 59mBig data beautiful visualization on the browser with BokehFabio Pliger · EuroPython 2015
• 36mMaking interactive visualizations easy, inside and outside of notebooksJulia Signell · PyData DC 2018
• 03h 07mData of Unusual Size: Interactive VisualizationPavithra Eswaramoorthy, Dharhas Pothina · PyCon US 2023
• 29mHoloViz: Visualization and Interactive Dashboards in PythonJean-Luc Stevens · PyCon Italia 2023
• 40mData-centric exploration using intake, dask, hvplot, and moreJulia Signell · PyData New York City 2019
• 37mVisualizing Large Datasets with JavaScript Using deck.glMarko Letic · RVA JavaScript Conf 2019