Data Access Workflow

Learn about the spatial data access workflow for the supported data sources.

The following figure shows the spatial vector data model and its internal data access workflow:

Description of the illustration spatial_ai_data_access_workflow.png

As seen in the preceding figure, creation of SpatialDataFrame depends on the data source:

• Spatial database: SpatialDataFrame is created by Spatial AI itself.
• Files: SpatialDataFrame is created through GeoPandas/Fiona/OGR.

You can convert a SpatialDataFrame to GeoPandas GeoDataFrame and its columns to NumPy arrays when more complex in-memory analysis is required or when the data needs to be passed to a standard library such as scikit-learn or PySAL. Conversely, you can also convert a GeoPandas GeoDataFrame to a SpatialDataFrame.

All Oracle Spatial AI API algorithms accept SpatialDataFrame as input. Some algorithms can also accept GeoPandas GeoDataFrame or NumPy arrays.