-
Notifications
You must be signed in to change notification settings - Fork 367
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Minimum Python 3.8 #1990
Minimum Python 3.8 #1990
Conversation
I just removed the last commit to bump the test tolerances. I'm hoping those are fixed by #1994 instead, which won't need a tolerance bump then. So, if that goes in, then the 3.9/3.10 test_examples should be fine here too. |
pyupgrade `find lib/cartopy -name "*.py"`
Python 3.8 will be the minimum supported version now, along with Numpy 1.19, which follows NEP29 support intervals. This also bumps other versions to a 2-year support cycle.
@@ -112,7 +111,7 @@ def test_global_scatter_wrap_no_transform(): | |||
@pytest.mark.natural_earth | |||
@pytest.mark.mpl_image_compare( | |||
filename='global_hexbin_wrap.png', | |||
tolerance=2 if MPL_VERSION < parse_version('3.2') else 0.5) | |||
tolerance=0.5) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The default is 0.5, so this can be dropped.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks, and apologies for the CI thrashing here. The conda-forge testing labels were not in all of the releases, so had to pin to a specific version and make sure matplotlib-base
is from the testing label too.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@greglucas Apologies on that. Probably those releases pre-date cartopy using them. It should be much easier keeping testing up-to-date nowadays.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
No problem at all, that actually wasn't too bad to diagnose compared to some of the other flaky failures we seem to have here ;)
Pin minimum CI to 3.2.1 because that has the test label at conda-forge which we need for the image tests. Remove other image comparison checks for less than 3.2.
This bumps the minimum supported version to Python 3.8 for the next release and also adds Python 3.10 to the tests.
https://numpy.org/neps/nep-0029-deprecation_policy.html