flatten geometry series (3D to 2D) in geopandas dataframe

geopandas_convert_geometry_3D_2D.py

# Often when reading in a ShapeFile from Basemap, you'll get: "ValueError: readshapefile can only handle 2D shape types"
# A trick can be to convert your geometry in your GeoPandas Dataframe and restoring the new flattened 2D geometry
# series back into a shapefile and try again.

# edit from http://stackoverflow.com/questions/33417764/basemap-readshapefile-valueerror  

from shapely.geometry import Polygon, MultiPolygon, shape, Point
import geopandas as gp
def convert_3D_2D(geometry):
    '''
    Takes a GeoSeries of 3D Multi/Polygons (has_z) and returns a list of 2D Multi/Polygons
    '''
    new_geo = []
    for p in geometry:
        if p.has_z:
            if p.geom_type == 'Polygon':
                lines = [xy[:2] for xy in list(p.exterior.coords)]
                new_p = Polygon(lines)
                new_geo.append(new_p)
            elif p.geom_type == 'MultiPolygon':
                new_multi_p = []
                for ap in p:
                    lines = [xy[:2] for xy in list(ap.exterior.coords)]
                    new_p = Polygon(lines)
                    new_multi_p.append(new_p)
                new_geo.append(MultiPolygon(new_multi_p))
    return new_geo

geodf_2d = gp.GeoDataFrame.from_file(shp_file) # plug_in your shapefile
geodf_2d.geometry = convert_3D_2D(geodf_2d.geometry) # new geodf with 2D geometry series

# geodf_2d.to_file(path + shapefile.shp, driver = 'ESRI Shapefile') will sore a shapefile with 2D shape types

If you're using shapely, here's another helpful hack to convert a geometry object from 3D to 2D:

geom = shapely.wkb.loads(
        shapely.wkb.dumps(geom, output_dimension=2))

@koshy1123 Thanks for the hack. It looks much cleaner

@koshy1123 Thanks too! Is also quite fast 😀

shapely.wkb.loads(shapely.wkb.dumps(geom, output_dimension=2)) is cleaner, but I think convert_3D_2D(geometry) is much faster if I am not mistaken

I think https://gis.stackexchange.com/questions/67210/convert-3d-wkt-to-2d-shapely-geometry/220374 may be fastest

amazing function thanks a lot

You saved my project. Thanks a lot!!!

Really helped me out - thank you :)

Thanks above, output_dimensions=2 ftw!

import geopandas as gpd
from shapely import wkb

_drop_z = lambda geom: wkb.loads(wkb.dumps(geom, output_dimension=2))
df.geometry = df.geometry.transform(_drop_z)

%timeit drop_z(df.geometry)                # 42.9 s ± 270 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)
%timeit drop_z_pygeos(df.geometry)         # 568 ms ± 12.7 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)
%timeit drop_z_convert_3D_2D(df.geometry)  # 43.6 s ± 541 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

detail

• The convert_3D_2D above only works for (Multi)Polygons.
• drop_z_convert_3D_2D is a rewrite of convert_3D_2D for all shapely geometry objects.
• drop_z_pygeos requires pygeos to be installed. at least until shapely 2 is released.
• drop_z_shapely_for uses a method suggested: shapely/shapely#709
• drop_z_shapely_transform is to compare between the transform and for method for looping between rows/geoms. I did these to explain why I changed the suggested method. >10% improvement.
• I ran this on LineStrings of UK rivers with 186,255 rows.

import shapely 
import geopandas as gpd

df = gpd.read_file(<Your Data Here>)


def convert_3D_2D(geometry):
  ''' Drop Z coordiates from GeoSeries, returning list if (Multi)Polygons
  Taken from:  https://gist.github.com/rmania/8c88377a5c902dfbc134795a7af538d8

  Takes a GeoSeries of 3D Multi/Polygons (has_z) and returns a list of 2D Multi/Polygons
  '''
  new_geo = []
  for p in geometry:
    if p.has_z:
      if p.geom_type == 'Polygon':
        lines = [xy[:2] for xy in list(p.exterior.coords)]
        new_p = Polygon(lines)
        new_geo.append(new_p)
      elif p.geom_type == 'MultiPolygon':
        new_multi_p = []
        for ap in p:
          lines = [xy[:2] for xy in list(ap.exterior.coords)]
          new_p = Polygon(lines)
          new_multi_p.append(new_p)
        new_geo.append(MultiPolygon(new_multi_p))
  return new_geo


def drop_z_shapelywkb(ds):
  ''' Drop Z coordinates from GeoSeries, returns GeoSeries
  Using this method:  https://gist.github.com/rmania/8c88377a5c902dfbc134795a7af538d8?permalink_comment_id=2893099#gistcomment-2893099
  '''
  _drop_z_wkb = lambda geom:  shapely.wkb.loads(shapely.wkb.dumps(geom, output_dimension=2))
  return ds.transform(_drop_z_wkb)


def drop_z_pygeos(ds):
  ''' Drop Z coordinates from GeoSeries, returns GeoSeries
  Requires pygeos to be installed, and such I've added `import pygeos` to check.
  '''
  import pygeos
  return gpd.GeoSeries.from_wkb(ds.to_wkb(output_dimension=2))


def drop_z_convert_3D_2D(ds):
  ''' Drop Z coordinates from GeoSeries, returns GeoSeries
  A rewrite of convert_3D_2D but to fit the form of these other drop_z... functions.
  '''
  _get_coords = lambda g:  (
    g.coords
    if not g.geom_type == 'Polygon' else  # isinstance(g, Polygon)?
    g.exterior.coords
  )
  _drop_z = lambda geom:  type(geom)(
    (_drop_z(g) for g in geom.geoms)  # nest for iterables
    if hasattr(geom, '__iter__') else
    (coord[:2] for coord in _get_coords(geom))  # keep x,y
  )
  return gpd.GeoSeries(_drop_z(geom) for geom in ds)


_drop_z_shapely = lambda geom:  shapely.ops.transform(lambda *args: args[:2], geom)


def drop_z_shapely_for(ds):
  ''' Drop Z coordinates from GeoSeries, returns GeoSeries
  Using this method: https://github.com/shapely/shapely/issues/709#issuecomment-799977173
  '''
  return gpd.GeoSeries(_drop_z_shapely(geom) for geom in ds)


def drop_z_shapely_transform(ds):
  ''' Drop Z coordinates from GeoSeries, returns GeoSeries
  Same as drop_z_shapely_for but using the pandas transform operation instead.
  '''
  return ds.transform(lambda geom: _drop_z_shapely(geom))



a = convert_3D_2D(df.geometry)             # [] (empty list because I have only LineStrings)
b = drop_z_shapelywkb(df.geometry)         # GeoSeries, 2D geoms
c = drop_z_pygeos(df.geometry)             # GeoSeries, 2D geoms
d = drop_z_convert_3D_2D(df.geometry)      # GeoSeries, 2D geoms
e = drop_z_shapely_for(df.geometry)        # GeoSeries, 2D geoms
f = drop_z_shapely_transform(df.geometry)  # GeoSeries, 2D geoms
print( all([
  gpd.GeoSeries.all(b==c),
  gpd.GeoSeries.all(c==d),
  gpd.GeoSeries.all(d==e),
  gpd.GeoSeries.all(e==f),
  gpd.GeoSeries.all(f==b),
]) )  # True



%timeit convert_3D_2D(df.geometry)             # 19.7 s ± 430 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)
%timeit drop_z_shapelywkb(df.geometry)         # 42.9 s ± 270 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)
%timeit drop_z_pygeos(df.geometry)             # 568 ms ± 12.7 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)
%timeit drop_z_convert_3D_2D(df.geometry)      # 43.6 s ± 541 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)
%timeit drop_z_shapely_for(df.geometry)        # 45.6 s ± 310 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)
%timeit drop_z_shapely_transform(df.geometry)  # 37.7 s ± 722 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

Thanks! it helped me get 2D Polygons from the KML file from Google Earth.

Thanks so much! This really helped me getting 2D polygons from Google Earth pro!

Is force_2d now the best approach for this?

Thanks! works great.