I have a large geographic dataset, including lots of points, that I'm creating an Mapbox vector tileset from using tippecanoe. Rather than dropping points, I'm clustering points. Separately, the original GeoJSON features have a unique id that I want to preserve in the tiles.
The trouble is showing up when I try to combine the two. I'm currently building tiles like this:
tippecanoe -o tiles/points.mbtiles -r1 -Z8 -z11 --cluster-distance=10 --use-attribute-for-id=unique_id geojson/points.geojson
The result is that all un-clustered features have the correct feature id (preserved from the original unique_id field). All well and good. But the resulting clustered features also have an id that comes from one of the features that make up the cluster.
Is it possible to tell tippecanoe to assign ids to clusters differently? Ideally, they'd be uuids.
I did try changing the options like this:
tippecanoe -o tiles/points.mbtiles -r1 -Z8 -z11 --cluster-distance=10 -y unique_id geojson/points.geojson
This just preserves unique_id in the features' properties objects, and still gives features ids in the same way as above.