Repository navigation
TST: Restore some tolerances for some arch/platform-specific failures - #31343
QuLogic merged 3 commits into
Conversation
Sorry, something went wrong.
This comment was marked as spam.
|
OK, this is now passing. The failure in test_use14corefonts does point out how backend-specific measurement is useful. PDF/PS support for using the core 14 fonts means it overrides get_text_width_height_descent and returns builtin metrics for those when the option is enabled. Since the line-height-from-font-metrics code doesn't use a backend-specific API for it, it doesn't see that override and is dependent on whether Helvetica is installed or not. So I regenerated it with Helvetica installed, but I've had to add a tolerance for other systems without it. |
Sorry, something went wrong.
|
OK, it does require a bit of internal trickery, but instead of the last commit, we can instead do #31371. |
Sorry, something went wrong.
When outputting files using the "core 14 fonts" (e.g., `rcParams['pdf.use14corefonts'] = True`), text will be measured using our internal AFM files instead of any external files. While we don't have a full API decided for how to pass full `Text` objects through backends, add in a small internal helper to allow the PS/PDF backends to override font height metrics with the AFM files.
|
Rebased on top of #31371 and dropped the AFM test change since that PR's approved. |
Sorry, something went wrong.
|
Test failures are a bit getting out of hand, so that I don't dare to merge. We should fix setuptools scm, likely temporarily by pinning <10.0, see #31387. |
Sorry, something went wrong.
|
OK, I've pinned setuptools-scm for this branch and I think I've got all the places. Unfortunately, since this branch hasn't been merged with main recently, it doesn't have #29281. |
Sorry, something went wrong.
|
OK, looks like I didn't increase one tolerance enough, and test_stairs_options is failing everywhere. I don't quite understand the latter yet as if I apply that image locally, it doesn't pass for me. |
Sorry, something went wrong.
So before the AFM change, the minimum ascent was 15.1953125, and after it's 15.195312499999996 which causes only the A in the legend to shift slightly. I don't think that's too important, so I've just updated the final figure in the other branch and restarted the CI here. |
Sorry, something went wrong.
|
Trying to re-trigger Azure, which seems to have not run lately. |
Sorry, something went wrong.
The tolerances on these were increased in various PRs, with notes to reduce/remove them, so they were removed in matplotlib#30184, but should have been reverted to a smaller value: - `test_floating_axes.py::test_curvelinear4[png]` fails on macOS - `test_grid_helper_curvelinear.py::test_axis_direction[png]` fails on ARM and macOS - `test_grid_helper_curvelinear.py::test_polar_box[png]` fails on Windows, ARM, and macOS These fail on macOS; there's tolerances on them already, they fail by just a small amount, so increase them: - `test_bbox_tight.py::test_bbox_inches_tight_suptile_legend[png]` - `test_patheffects.py::test_collection[png]` - `test_axes3d.py::test_scale3d_artists_log[png]` This fails only on AppVeyor (Windows), but not Azure. There is a tolerance there already, so just add it to the list: - `test_mathtext.py::test_mathtext_rendering`
This is currently causing warnings at runtime in the editable install, which breaks almost all tests.
|
Increased the tolerance for the test failure on AppVeyor, though it doesn't show on Azure. There's already a tolerance there, so just made it a bit bigger. |
Sorry, something went wrong.
PR summary
The tolerances on these were increased in various PRs, with notes to reduce/remove them, so they were removed in #30184, but should have been reverted to a smaller value:
These fail on macOS; there's tolerances on them already, they fail by just a small amount, so increase them:
This fails only on AppVeyor (Windows), but not Azure. There is a tolerance there already, so just add it to the list:
AI Disclosure
None
PR checklist