Repository navigation
Correctly pass location when constructing ImageGrid colorbar. - #25884
timhoffm merged 1 commit into
Conversation
Sorry, something went wrong.
Previously only the orientation would be passed, so information on whether the colorbar is on the top or the right would lost, resulting in ticks, ticklabels, and label being on the wrong side of the colorbar. Fix that. The baseline image changes exactly test that (the colorbar ticks move to the other side). While at it, also simplify the plotted data in test_imagegrid_cbar_mode, which decreases more than 2x the size of the baseline image (86K -> 34K).
| @@ -22,23 +22,14 @@ def __init__(self, *args, orientation, **kwargs): | |||
| self.orientation = orientation | |||
| super().__init__(*args, **kwargs) | |||
There was a problem hiding this comment.
Semi-OT, but got to this because you deleted cla(): This super().__init__ should not do anything, right? - Likely a leftover from the dedicated axes_grid colorbar classes?
Sorry, something went wrong.
There was a problem hiding this comment.
It does (should?), as CbarAxesBase is a mixin and will thus end up in the base classes of the actual colorbar axes class used by ImageGrid, and thus we need to forward arguments via super() (after extracting out orientation).
Sorry, something went wrong.
There was a problem hiding this comment.
Indeed. Maybe that should be mentioned somewhere in the class? without that context, it's looks like object is the only parent. Either way, the PR is correct.
Sorry, something went wrong.
Previously only the orientation would be passed, so information on whether the colorbar is on the top or the right would lost, resulting in ticks, ticklabels, and label being on the wrong side of the colorbar. Fix that.
The baseline image changes exactly test that (the colorbar ticks move to the other side). While at it, also simplify the plotted data in test_imagegrid_cbar_mode, which decreases more than 2x the size of the baseline image (86K -> 34K).
Closes #23595.
I think this was really broken by https://github.com/matplotlib/matplotlib/pull/20501/files#diff-ba79d31808d1f55bdf8bf2896366c85122eb9488e140c5a4de060f2c4e1fa8f6L42.
PR summary
PR checklist