Clustering in the Mapbox Tiling Service
summary
Clustering moves the heavy lifting into the cloud: MTS creates clustered and unclustered vector layers from line‑delimited GeoJSON, the demo uploads ~100k US public‑school points with the tilesets-cli, sets a region_count 256 for finer grouping, and publishes a tileset that returns precomputed clusters up to zoom 12 and raw points to zoom 16. The recipe shows features aggregation (sum count), min/max zoom rules, and styling examples (circle radii, heatmap); the pipeline runs end‑to‑end and the tileset renders as expected.
AI-written summary · about
chapters (15)
- 0:00Introduction to clustering
- 1:47What is Mapbox Tiling Service
- 3:32Why use clustering
- 7:29MTS vs client-side clustering
- 9:21Setting up the tile set CLI
- 12:30Authenticating with secret token
- 13:11Uploading source data
- 16:27Creating recipe configuration
- 20:22Understanding region count
- 23:56Creating and publishing tile set
- 26:31Visualizing clustered data
- 34:36Styling clusters with expressions
- 38:10Heat map clustering demo
- 39:57Optimizing large data sets
- 42:24Resources and Q&A
related talks
16mRaster tiles to vector tiles: how to cross the gulf
21mPoint-in-polygon lookups with vector tiles: quick, effective, and no server needed!
17mBoosting Map Performance with Mapbox and Flutter
15mHave you checked your vector tiles lately?
40mStronger App Architecture Using Maps
15mMaking maps work on mobile
01h 34mBuild a Web Map with Open Source Tools
32m"A Developer's Guide to Open Source Web Mapping Libraries"