I have a set of point geometries that I would like to cluster together using ST_ClusterDBSCAN. Unfortunately, it doesn't seem to work for WGS84 and requires a meter-based projection. Every tutorial that I can seem to find manually determines which projection is a good meter-based projection for their particular dataset. Is there a generalizable way to do this automatically?
Basically what I'd like to do is take the convex hull of my points, take the centroid of that polygon and then query to find what's a good meter-based projection for that particular region of the world. I'll then convert my points to this projection and run the clustering.