Work with spatial indexes
Use this guide when you need faster repeated spatial filtering or joins.
Current GeoDataFrames behavior
GeoDataFrames.read may store geometry columns in GeometryVector, but GeoDataFrames does not currently expose a public API to build and query a spatial index directly on that vector.
For join workflows, FlexiJoins performs candidate filtering with an STR tree on the right-side table before running the exact geometry predicate.
Speed up repeated spatial queries
For repeated predicate evaluation over larger datasets, use an index-oriented package and keep geometries in a table with one consistent CRS: 2. Reproject both datasets to a suitable projected CRS when distance or area assumptions matter.
Build or reuse the package's index structure once.
Use the index to narrow candidates, then apply exact predicates (
intersects,within,contains) to candidates.
See perform a spatial join for table-level join patterns and operations and joins background for design rationale.