Scatter3D: depthshade=False causes IndexError for Tkinter when plotting more than one point.

[OkkeVanEck]

Bug report

Bug summary

Setting depthshade=False causes an IndexError for Tkinter when plotting more than one point. There are no errors when plotting one point.

Code for reproduction

    fig = plt.figure()
    ax = fig.gca(projection="3d")
    ax.scatter([0,1], [0,1], [0,1], c="royalblue", marker="o",
                depthshade=False, s=60, label="H")
    plt.show()

Actual outcome

I'll only post the first and last rows of the terminal, as the rest contains privacy sensitive information.

First rows:

Exception in Tkinter callback
Traceback (most recent call last):
  File "/usr/lib/python3.6/tkinter/__init__.py", line 1705, in __call__
    return self.func(*args)
  File "/usr/lib/python3.6/tkinter/__init__.py", line 749, in callit
    func(*args)
...: line 253, in idle_draw
    self.draw()

Last rows:

.../venv/lib/python3.6/site-packages/mpl_toolkits/mplot3d/art3d.py", line 507, in do_3d_projection
    fcs = fcs[z_markers_idx]
IndexError: index 1 is out of bounds for axis 0 with size 1

Expected outcome

It should just plot two points in a 3D space.

Matplotlib version

• Operating system: Ubuntu 18.04
• Matplotlib version: 3.3.0
• Matplotlib backend TkAgg:
• Python version: 3.6.9
• Jupyter version (if applicable): -
• Other libraries:

I recently upgraded Matplotlib via pip:

pip install --upgrade matplotlib

And originally installed it with pip as well within a virtual environment created with python-venv. I tested the code above in an older virtual environment and no errors occur when using matplotlib version 3.2.2.

[tacaswell]

The full trackback:

Traceback (most recent call last):
  File "/home/tcaswell/.virtualenvs/bleeding/lib/python3.10/site-packages/matplotlib/backends/backend_qt5.py", line 480, in _draw_idle
    self.draw()
  File "/home/tcaswell/.virtualenvs/bleeding/lib/python3.10/site-packages/matplotlib/backends/backend_agg.py", line 407, in draw
    self.figure.draw(self.renderer)
  File "/home/tcaswell/.virtualenvs/bleeding/lib/python3.10/site-packages/matplotlib/artist.py", line 41, in draw_wrapper
    return draw(artist, renderer, *args, **kwargs)
  File "/home/tcaswell/.virtualenvs/bleeding/lib/python3.10/site-packages/matplotlib/figure.py", line 1863, in draw
    mimage._draw_list_compositing_images(
  File "/home/tcaswell/.virtualenvs/bleeding/lib/python3.10/site-packages/matplotlib/image.py", line 132, in _draw_list_compositing_images
    a.draw(renderer)
  File "/home/tcaswell/.virtualenvs/bleeding/lib/python3.10/site-packages/matplotlib/artist.py", line 41, in draw_wrapper
    return draw(artist, renderer, *args, **kwargs)
  File "/home/tcaswell/.virtualenvs/bleeding/lib/python3.10/site-packages/mpl_toolkits/mplot3d/axes3d.py", line 445, in draw
    sorted(self.collections,
  File "/home/tcaswell/.virtualenvs/bleeding/lib/python3.10/site-packages/mpl_toolkits/mplot3d/axes3d.py", line 446, in <lambda>
    key=lambda col: col.do_3d_projection(renderer),
  File "/home/tcaswell/.virtualenvs/bleeding/lib/python3.10/site-packages/mpl_toolkits/mplot3d/art3d.py", line 507, in do_3d_projection
    fcs = fcs[z_markers_idx]

This is a regression from 3.2.

[tacaswell]

Labeling as a good first issue because there is an example that used to work and no longer does so no API design work, but medium difficulty because the mplot3d code is a bit complex to walk through.

Tasks:

• track down bug (git bisect or read the code)
• fix bug + add test

[neil-and-void]

Hi, I'm new to open source, could I try my hand at this?

[QuLogic]

Have at it.

[neil-and-void]

4428da6 is the first bad commit (cc @xenoryt)

Author: Tyrone Xiong ty.cx123@gmail.com
Date: Wed Mar 14 19:44:57 2018 -0400

I may have come up with some type of solution but I don't know why it works due to the complex names of the variables within this section

[neil-and-void]

take

[QuLogic]

Fixed by #18156.