GeoPandas

If you are processing geospatial data that’s small enough to fit in memory, then besides (Databricks SQL and) DuckDB Spatial, you can also consider using GeoPandas. For more scalable approaches, you could define Spark UDFs based on GeoPandas similar to how we defined them on top of DuckDB Spatial. As this guide is mostly focused on SQL-based approaches, we won’t go into more detail here.