[FIXED] How remove residual lines after merge two regions using Geopandas in python?

Issue

import geopandas as gpd

sg = gpd.read_file('shapefile1.shp')
sd = gpd.read_file('shapefile2.shp')

kw = sg[sg['region_name'] == 'region1']
tg = sd[sd['region_name'] == 'region2']

region1 and region2 are as below:

mr = kw.geometry.reset_index(drop=True).union(tg.geometry.reset_index(drop=True))
mr.plot()

I can get the image from mr as below:

But, see the boundary of the region

mr.boundary.plot()

I can see the residual lines as the image above.

How remove these residual lines?

I'v tried to remove them using simplify(tolerance=0.01) and buffer(0) functions. But I failed.

I just expect to get outlines of the first image with empty inside.

Solution

For the benefits of the readers, I offer this answer that shows the relevant steps including runnable code that is missing in the question.

import os
os.environ['USE_PYGEOS'] = '0'
import geopandas as gpd
from shapely.geometry import Polygon

# Create 2 geodataframes each containing a simple polygon 
polygon_geom1 = Polygon(zip([0, 1.05, 1.05, 0], [-0.1, -0.1, 1.2, 1.2]))
#crs = {'init': 'epsg:4326'}
crs = 'epsg:4326'
polygon1 = gpd.GeoDataFrame(index=[0], crs=crs, geometry=[polygon_geom1])

polygon_geom2 = Polygon(zip([0+1, 1+1, 1+1, 0+1, 0.1+1, 0+1], [0,0,1,1,0.5,0]))
polygon2 = gpd.GeoDataFrame(index=[0], crs=crs, geometry=[polygon_geom2])

# Plot to see them together
ax1 = polygon1.plot(ec='k', color='y')
polygon2.plot(ax=ax1, ec='r', color='lightgray', alpha=0.8)

Next, we do the union of the two geometries:-

polygon1.geometry.values[0].union(polygon2.geometry.values[0])

then, the exterior geometry:-

polygon1.geometry.values[0].union(polygon2.geometry.values[0]).exterior

and, finally, the interior geometry:-

# The interiors of the unioned geometry
# This show item 0 of them
polygon1.geometry.values[0].union(polygon2.geometry.values[0]).interiors[0]

Part 2

This is another example with modified input geometries that will create 2 holes when perform "union" operation.

The code below will show how to create a union-polygon that keeps the large hole in it.

import os
os.environ['USE_PYGEOS'] = '0'
import geopandas as gpd
from shapely.geometry import Polygon

# Create 2 geodataframes each containing a simple polygon 
polygon_geom1 = Polygon(zip([0, 1.05, 1.05, 0], [-0.1, -0.1, 1.2, 1.2]))
#crs = {'init': 'epsg:4326'}
crs = 'epsg:4326'
polygon1 = gpd.GeoDataFrame(index=[0], crs=crs, geometry=[polygon_geom1])

polygon_geom2 = Polygon(zip([1, 2, 2, 1, 
                         1.1, 0.95, 1.2 ,
                         1], 
                        [0,0,1,1,
                         0.8,0.7,0.4,
                         0]))
polygon2 = gpd.GeoDataFrame(index=[0], crs=crs, geometry=[polygon_geom2])

# Plot to see them together
ax1 = polygon1.plot(ec='k', color='y')
polygon2.plot(ax=ax1, ec='r', color='lightgray', alpha=0.7)

# The union of the two geometries
# There are 2 holes in it
union_poly = polygon1.geometry.values[0].union(polygon2.geometry.values[0])
union_poly

# The exterior of the unioned geometry
ext_poly = polygon1.geometry.values[0].union(polygon2.geometry.values[0]).exterior
ext_poly

# The interiors of the unioned geometry
# This shows item 0 of them
int_poly0 =  polygon1.geometry.values[0].union(polygon2.geometry.values[0]).interiors[0]
int_poly0

# The polygon that contain the big hole only
# (small hole is excluded)
diff_poly = Polygon(ext_poly).difference(Polygon(int_poly0))
diff_poly

Answered By - swatchai