Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add functionality to drop 3rd dimension #709

LostFan123 opened this issue May 5, 2019 · 2 comments


Copy link

commented May 5, 2019

There is an old question on GIS Stack Exchange about converting 3D geometries to 2D: Convert 3D WKT to 2D Shapely Geometry. I think this functionality should be included in the Shapely, so we could use it, for example, like:

>>> from shapely.geometry import Polygon
>>> p = Polygon([(0, 0, 0), (1, 0, 0), (1, 1, 0)])
>>> p.wkt
'POLYGON Z ((0 0 0, 1 0 0, 1 1 0, 0 0 0))'
>>> p2 = p.drop_z
>>> p2.wkt
'POLYGON ((0 0, 1 0, 1 1, 0 0))'

I've seen in one of the answers and in the docs that this operation is not necessary as the 3rd dimension has no effect on geometric analysis. But when implementing my own function to determine left side of a split geometry (this is not yet implemented: #589), I saw that this functionality could come handy:

from shapely.geometry import (LinearRing, 

def is_left(polygon: Polygon,
            line: LineString) -> bool:
    Determines if the polygon is on the left side of the line
    according to:
    ring = LinearRing([*line.coords, *polygon.centroid.coords])
    return ring.is_ccw

This code will fail for 3D geometries:

p = Polygon([(0, 0, 0), (1, 0, 0), (1, 1, 0)])
l = LineString([(0, 0, 0), (1, 0, 0)])
is_left(p, l)

will give this error:

AttributeError                            Traceback (most recent call last)
~/miniconda3/lib/python3.7/site-packages/shapely/speedups/_speedups.pyx in shapely.speedups._speedups.geos_linearring_from_py()

AttributeError: 'list' object has no attribute '__array_interface__'

During handling of the above exception, another exception occurred:

IndexError                                Traceback (most recent call last)
<ipython-input-55-555b9e2533fa> in <module>()
----> 1 is_left(p, l)

<ipython-input-52-7fad75b19ce3> in is_left(polygon, line)
      7     """
----> 8     ring = LinearRing([*line.coords, *polygon.centroid.coords])
      9     return ring.is_ccw

~/miniconda3/lib/python3.7/site-packages/shapely/geometry/ in __init__(self, coordinates)
     51         BaseGeometry.__init__(self)
     52         if coordinates is not None:
---> 53             self._set_coords(coordinates)
     55     @property

~/miniconda3/lib/python3.7/site-packages/shapely/geometry/ in _set_coords(self, coordinates)
     66     def _set_coords(self, coordinates):
     67         self.empty()
---> 68         ret = geos_linearring_from_py(coordinates)
     69         if ret is not None:
     70             self._geom, self._ndim = ret

~/miniconda3/lib/python3.7/site-packages/shapely/speedups/_speedups.pyx in shapely.speedups._speedups.geos_linearring_from_py()

IndexError: tuple index out of range

This is due to the fact that centroid of a Polygon is always returned in 2D (#554), and a LinearRing can't be constructed from points having a different number of dimensions.

If there was a drop_z method, I would just write ring = LinearRing([*line.drop_z.coords, *polygon.centroid.coords]) instead of cluttering the code with things like line = LineString([xy[:2] for xy in list(line.coords)]) or implementing a function for that. Or even better, I would drop redundant 3rd dimension consisting only of zeros from the original parent polygon that I read from a file on the top level, so all the child geometries would have only 2 dimensions.

Shapely version: 1.6.4.post1, installed from conda.


This comment has been minimized.

Copy link

commented May 6, 2019

@LostFan123 I think a method to drop Z values would be an appropriate addition for 1.7. Thanks for suggesting it.

@sgillies sgillies added this to the 1.7 milestone May 6, 2019
@sgillies sgillies added the enhancement label May 6, 2019
@sgillies sgillies self-assigned this May 6, 2019

This comment has been minimized.

Copy link

commented Aug 23, 2019

For other people arriving here from Google, I found a very simple method:

def _to_2d(x, y, z):
    return tuple(filter(None, [x, y]))

new_shape = shapely.ops.transform(_to_2d, shape)

(credit to @feenster and @hunt3ri from ❤️)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
None yet
3 participants
You can’t perform that action at this time.