Add region selection from file

[acordonez]

This PR adds a function that takes an xarray dataset and returns the dataset selected over a region read from file. Grid cells outside this region are converted to NaN.

Geopandas read_file() is used to load the input file. While geopandas.read_file() can load many different file types, this has only been tested with a shapefile (.shp).

The regionmask library is used to extract the gridded data over the region of interest.

Example:

import xcdat
from pcmdi_metrics.io.region_from_file import region_from_file

# An xarray dataset to work with
ds = xc.open_dataset("my_test_dataset.nc")
fshp = "ne_110m_admin_0_countries/ne_110m_admin_0_countries.shp"
attr = "NAME_EN"
feature = "Tanzania"
ds_rg = region_from_file(ds,fshp,feature,attr)

[lee1043]

@acordonez thank you for the PR. Will we need to add geopandas as one of dependencies in dev.yml?

[acordonez]

@lee1043 It's probably a good idea since I'm calling it explicitly. Geopandas is already a dependency via regionmask and probably some of the other packages we use.

[lee1043]

@acordonez thanks for clarifying. Thanks for adding this capability. Please feel free to merge it!

[acordonez]

Thanks @lee1043!