pygeospandas is a simple wrapper around PyGEOS, in order to allow using its functionality in pandas.
It contains the following pieces:

  • GeosArray & GeosDtype
    pandas ExtensionArray and Dtype to be able to work with PyGEOS data in Series and DataFrames.

  • “geos” Series Accessor
    Access all of the PyGEOS functionality on Series objects, so that there is no need to unwrap to NumPy arrays.

  • “geos” DataFrame Accessor
    Apply PyGEOS functions to all “geos” Dtype series in the dataframe at once.

GeoPandas

This library is pretty similar to the GeoPandas library.
The main difference is that this library is much more lightweight, as it only depends on NumPy, pandas and PyGeos. You might want to use this library over GeoPandas if you need to work with geometries, but do not care about CRS or other geospatial features.

Note that this library only supports PyGEOS operations and is thus more limited in what it can do. There are functions to transform the data to Shapely geometries, which gives more possibilities, but only PyGEOS supported operations are implemented on the Series and DataFrames.

Indices and tables