a package to create and plot Voronoi regions in geographic areas
geovoronoi – a package to create and plot Voronoi regions inside geographic areas
Markus Konrad email@example.com, March 2020
geovoronoi helps generating Voronoi regions for geographic data, for example coordinates of public universities in a certain country. This in turn may be used to estimate some kind of "coverage".
It takes a list of coordinates and calculates the Voronoi regions from them using SciPy. At the edges, these regions go to infinity. We can then take the shape of the surrounding area (e.g. the shape of a country as polygon) to cut the Voronoi regions so that they fit into the provided shape, making the regions at the edges finite. geovoronoi uses shapely for these operations. The package furthermore implements some functions for easy plotting of the resulting Voronoi regions.
This package is available on PyPI. You can install the latest version via pip as follows:
# install with "plotting" dependencies (recommended): pip install -U geovoronoi[plotting] # or install base version: pip install -U geovoronoi
You have a geographic area that contains some points for which you want to generate Voronoi regions. This geographic area is a shapely Polygon/MultiPolygon object (that you, for example obtained from a GeoJSON file that you loaded with GeoPandas or Fiona). The N points you have are either in the form of a Nx2 NumPy array, or a list of shapely Point objects (they can be converted with the functions
points_to_coords). Both the points and the surrounding geographic area must have the same CRS (coordinate reference system).
Let's take for example these randomly generated points in Italy (in World Mercator CRS):
import numpy as np # coords = ... generate some coordinates with np.random.uniform ... print(coords)
array([[1690891.43454513, 4865911.53550427], [1303898.2749075 , 5398659.4816214 ], [1379407.32051822, 5701267.51923313], [1703402.05850744, 4916559.63783754], ... ]])
The surrounding shape of Italy was obtained beforehand from GeoPandas:
import geopandas as gpd world = gpd.read_file(gpd.datasets.get_path('naturalearth_lowres')) area = world[world.name == 'Italy'] area = area.to_crs(epsg=3395) # convert to World Mercator CRS area_shape = area.iloc.geometry # get the Polygon
Now we can calculate the Voronoi regions, cut them with the geographic area shape and assign the points to them:
from geovoronoi import voronoi_regions_from_coords poly_shapes, pts, poly_to_pt_assignments = voronoi_regions_from_coords(coords, area_shape)
poly_shapes is a list of shapely Polygon objects that represent the shape of the respective Voronoi region. With them, you can do everything that the shapely API provides. You can, for example, get each Voronoi region's area (there's also a helper function
calculate_polygon_areas in geovoronoi for that).
pts is list of shapely Point objects converted from
coords (in the same order).
poly_to_pt_assignments is a nested list that for each Voronoi region in
poly_shapes contains a list of indices into
pts (and hence
coords) that represent the points that belong to this Voronoi region. Usually, this is only a single point. However, in case of duplicate points (e.g. both or more points have exactly the same coordinates) then all these duplicate points are listed for the respective Voronoi region.
You can plot the results with the functions from the
import matplotlib.pyplot as plt from geovoronoi.plotting import subplot_for_map, plot_voronoi_polys_with_points_in_area fig, ax = subplot_for_map() plot_voronoi_polys_with_points_in_area(ax, area_shape, poly_shapes, coords, poly_to_pt_assignments) plt.show()
This would be an example output (see "Limitations" below for the issue of the regions in Sardinia):
See the full example source code in examples/random_points_across_italy.py. See also the other examples in the examples/ directory that show how to calculate the area of the Voronoi regions, handle duplicate points or interact with GeoPandas.
geovoronoi requires Python 3.6 or newer. The following packages need to be installed (if not, they will be automatically installed if you use a Python package manager like pip):
- matplotlib (only necessary for plotting)
- geopandas (only necessary for plotting)
- descartes (only necessary for plotting)
- MultiPolygon shapes of geographic areas are not treated separately when cutting the Voronoi regions (see Sardinia in
- support of MultiPolygon shapes of geographic areas (see Sardinia in
- support for plotting
area_shapeobjects with multiple geometries
Licensed under Apache License 2.0. See
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
|Filename, size||File type||Python version||Upload date||Hashes|
|Filename, size geovoronoi-0.2.0-py3-none-any.whl (17.1 kB)||File type Wheel||Python version py3||Upload date||Hashes View|
|Filename, size geovoronoi-0.2.0.tar.gz (14.3 kB)||File type Source||Python version None||Upload date||Hashes View|
Hashes for geovoronoi-0.2.0-py3-none-any.whl