search

oturns / geosnap

The Geospatial Neighborhood Analysis Package
https://oturns.github.io/geosnap-guide
BSD 3-Clause "New" or "Revised" License
243 stars 32 forks source link

update CI workflow #350

Closed jGaboardi closed 11 months ago

jGaboardi commented 1 year ago

This PR:

The goal of this PR is to get CI green again with the current matrix. In subsequent PRs we can prune down OS runs, then see about getting testing up and running for Python 3.11. However, numba is still not in available for 3.11 so that is another hurdle.

knaaptime commented 1 year ago

thanks for taking a look. On the mpl front, i think the blocker is proplot. Unpinning mpl will pass because we're not currently testing the plotting funcs, which should also change.

On one hand, I think the proplot fix is probably straightforward, so i'll take a look and see if i can send a PR the dev will merge, but taking a look at the issues/history, its not clear the project will be maintained much, so might also be time to refactor and remove. I like proplots aesthetics and it makes some stuff a lot easier, but its not absolutely critical

knaaptime commented 1 year ago

if/until we decide to refactor proplot out, we need to keep the matlotlib pin (because the pinning is done incorrectly in proplot)

jGaboardi commented 11 months ago

@knaaptime This is now fixed in geosnap, correct? We can close if so.

knaaptime commented 11 months ago

yeah, i made proplot optional while its in flux, which was a bit easier than i thought (but will make it a bit harder to make nice figs).

you probably got a billion emails today as i figured this out, but one thing i had to do differently here than the pysal stack is use coverage run -m pytest otherwise it would refuse to see any lines as run (despite all the tests running through)