split regions

[mathause]

Assume I want to split a country/ region into two regions - how can I achieve this? E.g. define an eastern and western US.

[mathause]

I don't think I will ever include something like this but let's add this here for reference:

import geopandas as gp
import pandas as pd
import regionmask
import shapely

countries = regionmask.defined_regions.natural_earth.countries_110

# convert regions to a gp.GeoDataFrame
data = dict(
        numbers=countries.numbers,
        abbrevs=countries.abbrevs,
        names=countries.names,
        geometry=countries.polygons,
    )
df = gp.GeoDataFrame.from_dict(data)


# create two polygons encompassing the eastern and western half of the US
lon = -98.24
p1 = shapely.geometry.Polygon([[-180, 0], [lon, 0], [lon, 90], [-180, 90]])
p2 = shapely.geometry.Polygon([[0, 0], [lon, 0], [lon, 90], [0, 90]])
df_polys = gp.GeoDataFrame(geometry=[p1, p2])

# select US
sel = df.abbrevs == "US"

# split in half
US_split = gp.overlay(df.loc[sel], df_polys, how='intersection')

# update the names & abbrevs
US_split.abbrevs = US_split.abbrevs.str.cat(["w", "e"])
US_split.names = US_split.names.str.cat(["west", "east"], sep=" ")
US_split.numbers =  [177, 178]

df_full = pd.concat([df.loc[~sel], US_split])

# create new regions
countries_us2 = regionmask.Regions(
        df_full["geometry"],
        numbers=df_full.numbers,
        names=df_full.names,
        abbrevs=df_full.abbrevs
        )

and a plot:

countries_us2.plot()