GeoDataFrame.from_dict is currently an unimplemented GeoPandas compatibility constructor in Sedona.\n\nThe implementation should:\n\n- delegate dictionary parsing, orientation, and column selection to GeoPandas;\n- preserve active geometry selection, CRS metadata, index and row order, empty inputs, and multiple geometry columns;\n- handle Spark schema inference when geometry-valued columns begin with missing values;\n- document that the dictionary is materialized on the driver and that large datasets should use a distributed Spark or Sedona data source instead; and\n- add focused tests across supported Spark, GeoPandas, and Shapely versions.\n\nThis is intentionally a driver-local constructor for small in-memory dictionaries, not a distributed ingestion API.
GeoDataFrame.from_dict is currently an unimplemented GeoPandas compatibility constructor in Sedona.\n\nThe implementation should:\n\n- delegate dictionary parsing, orientation, and column selection to GeoPandas;\n- preserve active geometry selection, CRS metadata, index and row order, empty inputs, and multiple geometry columns;\n- handle Spark schema inference when geometry-valued columns begin with missing values;\n- document that the dictionary is materialized on the driver and that large datasets should use a distributed Spark or Sedona data source instead; and\n- add focused tests across supported Spark, GeoPandas, and Shapely versions.\n\nThis is intentionally a driver-local constructor for small in-memory dictionaries, not a distributed ingestion API.