Mar 5, 2020
diff --git a/lib/mpl_toolkits/mplot3d/art3d.py b/lib/mpl_toolkits/mplot3d/art3d.py
    def get_vector(self, segments3d):
        """Optimize points for projection."""
        if len(segments3d):
            xs, ys, zs = np.row_stack(segments3d).T
            # row_stack would split and re-concatenate all rows, and so it's
            # much faster to reshape if the data is already in the right format
            if isinstance(segments3d, np.ndarray):
                segments3d_t = segments3d.reshape(-1, segments3d.shape[-1]).T
            else:
                segments3d_t = np.row_stack(segments3d).T
            ones = np.broadcast_to(np.ones((1,), dtype=segments3d_t.dtype),
                                   shape=(1, segments3d_t.shape[-1],))
            self._vec = np.concatenate((segments3d_t, ones), axis=0)
        else:  # row_stack can't stack zero arrays.
            xs, ys, zs = [], [], []
        ones = np.ones(len(xs))
        self._vec = np.array([xs, ys, zs, ones])

        indices = [0, *np.cumsum([len(segment) for segment in segments3d])]
            self._vec = np.ndarray(shape=(0, 4))

        # range is much faster than np.cumsum with a list argument, so we
        # use it whenever the faces have a constant number of vertices.
        if isinstance(segments3d, np.ndarray):
            face_vertices = segments3d.shape[1]
            segment_ends = range(face_vertices,
                                 face_vertices * len(segments3d) + 1,
                                 face_vertices)
        else:
            segment_ends = np.cumsum([len(segment) for segment in segments3d])
        indices = [0, *segment_ends]
        self._segslices = [*map(slice, indices[:-1], indices[1:])]

    def set_verts(self, verts, closed=True):
Original file line number	Diff line number	Diff line change
Expand Up		@@ -578,13 +578,28 @@ def set_zsort(self, zsort):
		def get_vector(self, segments3d):
		"""Optimize points for projection."""
		if len(segments3d):
		xs, ys, zs = np.row_stack(segments3d).T
		# row_stack would split and re-concatenate all rows, and so it's
		# much faster to reshape if the data is already in the right format
		if isinstance(segments3d, np.ndarray):
		segments3d_t = segments3d.reshape(-1, segments3d.shape[-1]).T
		else:
		segments3d_t = np.row_stack(segments3d).T
		ones = np.broadcast_to(np.ones((1,), dtype=segments3d_t.dtype),
		shape=(1, segments3d_t.shape[-1],))
		self._vec = np.concatenate((segments3d_t, ones), axis=0)
		else: # row_stack can't stack zero arrays.
		xs, ys, zs = [], [], []
		ones = np.ones(len(xs))
		self._vec = np.array([xs, ys, zs, ones])

		indices = [0, *np.cumsum([len(segment) for segment in segments3d])]
		self._vec = np.ndarray(shape=(0, 4))

		# range is much faster than np.cumsum with a list argument, so we
		# use it whenever the faces have a constant number of vertices.
		if isinstance(segments3d, np.ndarray):
		face_vertices = segments3d.shape[1]
		segment_ends = range(face_vertices,
		face_vertices * len(segments3d) + 1,
		face_vertices)
		else:
		segment_ends = np.cumsum([len(segment) for segment in segments3d])
		indices = [0, *segment_ends]
		self._segslices = [*map(slice, indices[:-1], indices[1:])]

		def set_verts(self, verts, closed=True):
Expand Down