Fix display tests for matplotlib 3.9 #383
Milestone
Comments
|
Had a chance to test this locally, it's a grab bag. Most failures are indeed due to legend location, and we can fix this by setting the loc explicitly. Segment displays are somehow back to not setting limits properly though, so the viewport is completely wrong. These will take some work to fix. |
|
Ok, I think this issue with segments is due to this change: https://matplotlib.org/stable/api/prev_api_changes/api_changes_3.9.0.html#boxplots-now-ignore-masked-data-points - axvspan now creates Rectangle instead of Polygon patches, and for some reason that means they are now not participating in data limit calculations. |
|
Issue punted upstream to matplotlib/matplotlib#28383 |
|
Looks like the fix is merged and will be in the 3.9.1 release. |
guillaumekh
added a commit
to guillaumekh/jams
that referenced
this issue
Jul 2, 2024
guillaumekh
added a commit
to guillaumekh/jams
that referenced
this issue
Jul 4, 2024
note: - [email protected] dropped support for Python <= 3.8 - [email protected]+ broke mir_eval. A fix was merged in mir_eval@main but has not been released yet - see: mir-evaluation/mir_eval#352 - [email protected] brakes mir_eval - see: mir-evaluation/mir_eval#383 - i updated pytest hoping it would help solve failing tests but that's a HUGE v3—>v8 jump. I only made the bare minimum of fixes to accomodate this upgrade, and I would expect more work is needed - i wanted to add python 11/12 support but delayed this until we can already get tests runnning properly on 3.9/3.10
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
This came up in #382 , but matplotlib 3.9 is causing some regressions on display tests. From a quick read-through of the changelog, I don't see any obvious culprits.
It may have something to do with legend positioning or tick markers. 🤷
The text was updated successfully, but these errors were encountered: