[FIXED] GeoPandas .clip() on example from the docs throws TypeError

Issue

I am unable to reproduce the .clip()-example from the GeoPandas docs without error. I suspect it has something to do with my setup, since the same thing worked a few months ago in a different environment and I have not found reports of this happening to others. But I cannot figure out what the problem is--I am hoping someone here has an idea.

Copying and pasting the example code into my jupyter notebook looks something like this:

import geopandas
from shapely.geometry import Polygon

# get a set of points
capitals = geopandas.read_file(geopandas.datasets.get_path("naturalearth_cities"))

# Create a custom polygon
polygon = Polygon([(0, 0), (0, 90), (180, 90), (180, 0), (0, 0)])

# Attempt to clip points by polygon
capitals_clipped = capitals.clip(polygon)

Running it gives me the following error:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
Input In [7], in <cell line: 1>()
----> 1 capitals_clipped = capitals.clip(polygon)

File ~/.conda/envs/test-env/lib/python3.8/site-packages/pandas/util/_decorators.py:311, in deprecate_nonkeyword_arguments.<locals>.decorate.<locals>.wrapper(*args, **kwargs)
    305 if len(args) > num_allow_args:
    306     warnings.warn(
    307         msg.format(arguments=arguments),
    308         FutureWarning,
    309         stacklevel=stacklevel,
    310     )
--> 311 return func(*args, **kwargs)

File ~/.conda/envs/test-env/lib/python3.8/site-packages/pandas/core/frame.py:10917, in DataFrame.clip(self, lower, upper, axis, inplace, *args, **kwargs)
  10905 @deprecate_nonkeyword_arguments(
  10906     version=None, allowed_args=["self", "lower", "upper"]
  10907 )
   (...)
  10915     **kwargs,
  10916 ) -> DataFrame | None:
> 10917     return super().clip(lower, upper, axis, inplace, *args, **kwargs)

File ~/.conda/envs/test-env/lib/python3.8/site-packages/pandas/core/generic.py:7569, in NDFrame.clip(self, lower, upper, axis, inplace, *args, **kwargs)
   7567 result = self
   7568 if lower is not None:
-> 7569     result = result._clip_with_one_bound(
   7570         lower, method=self.ge, axis=axis, inplace=inplace
   7571     )
   7572 if upper is not None:
   7573     if inplace:

File ~/.conda/envs/test-env/lib/python3.8/site-packages/pandas/core/generic.py:7417, in NDFrame._clip_with_one_bound(self, threshold, method, axis, inplace)
   7414 else:
   7415     threshold_inf = threshold
-> 7417 subset = method(threshold_inf, axis=axis) | isna(self)
   7419 # GH 40420
   7420 return self.where(subset, threshold, axis=axis, inplace=inplace)

File ~/.conda/envs/test-env/lib/python3.8/site-packages/pandas/core/ops/__init__.py:470, in flex_comp_method_FRAME.<locals>.f(self, other, axis, level)
    466 axis = self._get_axis_number(axis) if axis is not None else 1
    468 self, other = align_method_FRAME(self, other, axis, flex=True, level=level)
--> 470 new_data = self._dispatch_frame_op(other, op, axis=axis)
    471 return self._construct_result(new_data)

File ~/.conda/envs/test-env/lib/python3.8/site-packages/pandas/core/frame.py:6973, in DataFrame._dispatch_frame_op(self, right, func, axis)
   6970 if not is_list_like(right):
   6971     # i.e. scalar, faster than checking np.ndim(right) == 0
   6972     with np.errstate(all="ignore"):
-> 6973         bm = self._mgr.apply(array_op, right=right)
   6974     return self._constructor(bm)
   6976 elif isinstance(right, DataFrame):

File ~/.conda/envs/test-env/lib/python3.8/site-packages/pandas/core/internals/managers.py:302, in BaseBlockManager.apply(self, f, align_keys, ignore_failures, **kwargs)
    300 try:
    301     if callable(f):
--> 302         applied = b.apply(f, **kwargs)
    303     else:
    304         applied = getattr(b, f)(**kwargs)

File ~/.conda/envs/test-env/lib/python3.8/site-packages/pandas/core/internals/blocks.py:402, in Block.apply(self, func, **kwargs)
    396 @final
    397 def apply(self, func, **kwargs) -> list[Block]:
    398     """
    399     apply the function to my values; return a block if we are not
    400     one
    401     """
--> 402     result = func(self.values, **kwargs)
    404     return self._split_op_result(result)

File ~/.conda/envs/test-env/lib/python3.8/site-packages/pandas/core/ops/array_ops.py:283, in comparison_op(left, right, op)
    280     return invalid_comparison(lvalues, rvalues, op)
    282 elif is_object_dtype(lvalues.dtype) or isinstance(rvalues, str):
--> 283     res_values = comp_method_OBJECT_ARRAY(op, lvalues, rvalues)
    285 else:
    286     res_values = _na_arithmetic_op(lvalues, rvalues, op, is_cmp=True)

File ~/.conda/envs/test-env/lib/python3.8/site-packages/pandas/core/ops/array_ops.py:73, in comp_method_OBJECT_ARRAY(op, x, y)
     71     result = libops.vec_compare(x.ravel(), y.ravel(), op)
     72 else:
---> 73     result = libops.scalar_compare(x.ravel(), y, op)
     74 return result.reshape(x.shape)

File ~/.conda/envs/test-env/lib/python3.8/site-packages/pandas/_libs/ops.pyx:107, in pandas._libs.ops.scalar_compare()

TypeError: '>=' not supported between instances of 'str' and 'Polygon'

So far I have tried to repeat this in a clean conda environment obtained like so:

conda create -n test-env
conda activate test-env
conda install ipykernel geopandas
ipython kernel install --user --name=test_ipython
conda deactivate test-env

According to conda, running python 3.7.11 and geopandas 0.9.0. Even using this fresh test_ipython-Kernel, I get the same TypeError when I attempt clipping the stock world map.

I don't currently think this is a bug, and am assuming mere ignorance on my part.

Solution

This was a problem with geopandas versions.

It turns out that geopandas.GeoDataFrame.clip() did not work the same way in v0.9.0. Checking the docs for the appropriate version of geopandas reveals that back then, clip was not a GeoDataFrame-method but a standalone one, making the solution simply

capitals_clipped = geopandas.clip(capitals, polygon)

Answered By - maitagorri