Fix IndexError when using scatter3d and depthshade=False #18156
Conversation
| @mpl3d_image_comparison(['scatter3d_depthshade_false.png']) | ||
| def test_scatter3d_depthshade_false(): | ||
| """ | ||
| Test that 3d scatter plot works with depthshade=False (issue #18037) |
There was a problem hiding this comment.
We try to be stingy with our image tests. If this is just a test to see if depthshade=False works then maybe if doesn't need the image comparison (faster, takes less disk space etc)?
There was a problem hiding this comment.
I'll work on changing this. Thanks for the feedback!
There was a problem hiding this comment.
... you are allowed to claim it does need an image test, but no need to include one if its not necessary...
There was a problem hiding this comment.
I think this might need an image test. I can't think of another way to generate a baseline figure without using the method being tested
There was a problem hiding this comment.
You can always generate a figure, but not check the image output (just don't add the mpl3d_image_comparison decorator). Since the original bug resulted in a traceback, that would be enough to test that the bug is fixed.
There was a problem hiding this comment.
@neilZon You don't need to compare figures either. You can just do
x = y = z = np.arange(16)
ax_test = fig_test.add_subplot(projection='3d')
ax_test.scatter(x, y, z, depthshade=False)and the figure will get made. If the code you are testing throws an error, the test will fail. If not, then it will pass. We call this a "smoketest" because it just checks that the code runs, and nothing else. But bear in mind this test would fail on master right now, so its still a worthwhile (and quick!) test.
There was a problem hiding this comment.
Oh ok, I understand now. Thank you for the guidance!
There was a problem hiding this comment.
I can't seem to get my code to fail now with this test
|
This looks fine to me. And thanks for adding a test, though see the note above! |
There was a problem hiding this comment.
Thanks for the PR! The fix looks good to me. For the test, please could you remove the test image, and I don't think a figure comparison needs to be done. It is fine just to check that the code runs without producing an error, so the check_figures_equal can be removed, and you only have to create one figure/axes to check the code works.
|
Squash merging to make sure the image doesn't get committed |
… depthshade=False
…156-on-v3.3.x Backport PR #18156 on branch v3.3.x (Fix IndexError when using scatter3d and depthshade=False)
PR Summary
Minor bug fix for issue #18037 that resulted in an IndexError when plotting more than 2 points on 3d scatter plot with depthshade=False
PR Checklist