Bug summary

The initialization of an mpl_toolkits.mplot3d.Poly3DCollection object cannot properly handle the parameter verts when verts = a list of (N, 3) array-like nested tuples, and shade=False.

Code for reproduction

from mpl_toolkits.mplot3d import art3d
corners = ((0, 0, 0), (0, 5, 0), (5, 5, 0), (5, 0, 0))
tri = art3d.Poly3DCollection([corners], shade=True) # Failed when shade=True
# tri = art3d.Poly3DCollection([corners]) # Passed with the default setting shade=False

Actual outcome

TypeError Traceback (most recent call last)
Cell In[3], line 1
----> 1 tri = art3d.Poly3DCollection([corners], shade=True)

File ~/anaconda3/envs/testmpl/lib/python3.12/site-packages/mpl_toolkits/mplot3d/art3d.py:905, in Poly3DCollection.init(self, verts, zsort, shade, lightsource, *args, **kwargs)
875 """
876 Parameters
877 ----------
(...)
902 and _edgecolors properties.
903 """
904 if shade:
--> 905 normals = _generate_normals(verts)
906 facecolors = kwargs.get('facecolors', None)
907 if facecolors is not None:

File ~/anaconda3/envs/testmpl/lib/python3.12/site-packages/mpl_toolkits/mplot3d/art3d.py:1222, in _generate_normals(polygons)
1220 n = len(ps)
1221 i1, i2, i3 = 0, n//3, 2*n//3
-> 1222 v1[poly_i, :] = ps[i1, :] - ps[i2, :]
1223 v2[poly_i, :] = ps[i2, :] - ps[i3, :]
1224 return np.cross(v1, v2)

TypeError: tuple indices must be integers or slices, not tuple

Expected outcome

No error.

Additional information

When shade=True, the __init__ function will first call the function _generate_normals(polygons), where polygons=verts. In our case, verts is not an instance of np.ndarray but a list, so it enters the for loop:

        for poly_i, ps in enumerate(polygons):
            n = len(ps)
            i1, i2, i3 = 0, n//3, 2*n//3
            v1[poly_i, :] = ps[i1, :] - ps[i2, :]
            v2[poly_i, :] = ps[i2, :] - ps[i3, :]

polygons is [((0, 0, 0), (0, 5, 0), (5, 5, 0), (5, 0, 0))], and ps is a nested tuple, so we need to convert ps to np.ndarray before using the array slicing like ps[i1, :]. A possible fix may be setting ps = np.asarray(ps) before array slicing.

Operating system

No response

Matplotlib Version

3.9.2

Matplotlib Backend

No response

Python version

No response

Jupyter version

No response

Installation

None