Using Julia & D3 to analyse web performance data

on YouTube ↗47m

" summary

Uses Julia and D3 to parse Boomerang web timing data, serialize rowwise JSON from DataFrames, and render per-browser histograms in iframes. A prototype with 250,000 circles hits DOM limits, switching to canvas becomes 100x faster. Buckets come from an IQR-based thresholder (about 25 interior bins plus outliers). Data flows via URL/postMessage or websocket (interact.JL) from Julia box to D3; multiple histograms are generated and compare distributions successfully.

AI-written summary · about

" chapters (5)

1. 0:00Start
2. 2:01GET SETUP
3. 10:34BASIC JULIA TUTORIAL
4. 38:27BASIC D3 TUTORIAL
5. 46:58INTERACT.JL

" related talks

• 33mPainting with DataRichard Westenra · RS Conf 2019
• 51mD3 js: Easy and flexible data visualisation using Web standardsJos Dirksen · Devoxx Belgium 2017
• 12mBattle-Tested Techniques for Animation & Data VizKrystal Campioni · React Summit Remote Edition 2021
• 46mD3 + Websocket for Live Web ApplicationsLoïc Hoguin · Erlang Factory SF Bay Area 2014
• 28mBuilding apps with D3.jsNathan Vander Wilt · CascadiaJS 2013
• 01h 55mUse your data to tell storiesAstrid Belder, Charalampos Xanthopoulakis · Devoxx Belgium 2019
• 24mBeyond the bar graph: Creative visualization with D3.jsJoanne Cheng · FutureJS 2014
• 59mBig data beautiful visualization on the browser with BokehFabio Pliger · EuroPython 2015