Clustering in the Mapbox Tiling Service

Clustering Point data within a vector tileset is a critical UI/UX improvement when large or dense datasets are displayed to a user. In this session, Andrew Sepic, Frontend Engineer on the Mapbox Documentation & Developer Tools team, explains the Mapbox Tiling Service (MTS), how to configure MTS recipes to support clustering, and how to style clusters using Mapbox Studio. Highlights include: - Learn how MTS is a tool for creating vector and raster tilesets by transforming data with "recipes." - Discover how clustering groups geographic points into a single feature at lower zoom levels, ideal for visualizing large point datasets. - Compare client-side clustering with MTS clustering, emphasizing the performance benefits of MTS for large datasets. - A live demonstration of how to create a clustered tileset using the tileset CLI, including authenticating with a secret token, uploading source data, defining a recipe.json file, creating the tileset, and publishing it. The recipe includes setting minimum and maximum zoom levels for clustered and unclustered data, defining features and attributes, and configuring clustering parameters like region count. - Tips for optimizing large datasets, such as keeping layer size within limits and removing unnecessary feature properties. Session links: Clustering in GL JS: https://docs.mapbox.com/mapbox-gl-js/example/cluster/ https://docs.mapbox.com/api/maps/mapbox-tiling-service/ https://github.com/mapbox/tilesets-cli https://www.mapbox.com/blog/data-clustering-now-available-in-the-mapbox-tiling-service More about over zoom: https://docs.mapbox.com/help/glossary/overzoom/ And considering it in MTS tileset design: https://docs.mapbox.com/help/troubleshooting/multilayer-tilesets/#consider-the-overzooming-behavior https://docs.mapbox.com/help/troubleshooting/adjust-tileset-zoom-extent/#minimum-and-maximum-zoom-levels