Raster Data I/O

The standard mechanism by which any data is brought in and out of a Spark Dataframe is the Spark SQL DataSource. RasterFrames provides specialized DataSources for geospatial raster data and maintains compatibility with existing general purpose DataSources, such as Parquet.

Furthermore, when in a Jupyter Notebook environment, you can view Tile and DataFrame samples.

There is also support for vector data for masking and data labeling.