COMPAT: mpl 3.0

[TomAugspurger]

Closes #22790

[pep8speaks]

Hello @TomAugspurger! Thanks for submitting the PR.

• There are no PEP8 issues in the file pandas/plotting/_compat.py !

• There are no PEP8 issues in the file pandas/plotting/_core.py !

• There are no PEP8 issues in the file pandas/tests/plotting/common.py !

• There are no PEP8 issues in the file pandas/tests/plotting/test_datetimelike.py !

[TomAugspurger]

The test pandas/tests/plotting/test_datetimelike.py::TestTSPlot::test_high_freq takes 123s for me locally. That seems excessive. I wonder if there was a performance regression there, or whether we can make the test smaller.

[TomAugspurger]

5fbffe4 was the original commit (6 years ago). Nothing in that looks like it depends on the number of points, just the freq, so I'm going to reduce the number to 100 (finishes in 0.5s locally with that).

[jreback]

could xfail?

[TomAugspurger]

This is in the plotting code itself, not in the test. The workaround is necessary for mpl < 3.0, but breaks things in mpl >= 3.0

[TomAugspurger]

@jreback good to merge on green, to get master passing again? (aside from #22871 which is failing our strict warnings build)

[jreback]

yep

[codecov]

Codecov Report

Merging #22870 into master will decrease coverage by <.01%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master   #22870      +/-   ##
==========================================
- Coverage   92.19%   92.18%   -0.01%     
==========================================
  Files         169      169              
  Lines       50827    50830       +3     
==========================================
  Hits        46860    46860              
- Misses       3967     3970       +3

Flag	Coverage Δ
#multiple	90.6% <100%> (-0.01%)	⬇️
#single	42.37% <33.33%> (-0.01%)	⬇️

Impacted Files	Coverage Δ
pandas/plotting/_compat.py	91.3% <100%> (+0.39%)	⬆️
pandas/plotting/_core.py	83.35% <100%> (-0.2%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update d115900...2ca4963. Read the comment docs.

[gfyoung]

Given that 0.23.x is still active, I would recommend backporting CI fixes like this. When I attempted to backport an actual regression last time, I ran into numerous CI-related issues unrelated to the original fix.