search

pysal / segregation

Segregation Measurement, Inferential Statistics, and Decomposition Analysis
https://pysal.org/segregation/
BSD 3-Clause "New" or "Revised" License
111 stars 26 forks source link

remove pygeos #210

Closed martinfleis closed 1 year ago

martinfleis commented 1 year ago

Not sure why pygeos is in requirements and CI ends here as it doesn't seem to be used in the code. Removing.

martinfleis commented 1 year ago

Huh? How come this PR caused a failure? Anyone closer to segregation code can tell me what is wrong?

knaaptime commented 1 year ago

weird. I'll dig into it

thanks for putting this together

jGaboardi commented 1 year ago

Very strange indeed. Seems to be coming from deep within joblib/futures?

jGaboardi commented 1 year ago

How and why removing the pygeos dependency would lead to that path is a bit of mystery...

martinfleis commented 1 year ago

It is not caused by this PR. See main https://github.com/pysal/segregation/actions/runs/4642225334/jobs/8215907349. For some reason, the workflow is marked as successful even when it failed in all envs :D.

martinfleis commented 1 year ago

https://github.com/pysal/segregation/blob/b3643f7af0aeaed1caf5a39b4e57c19a5c5f3459/.github/workflows/unittests.yml#L16

knaaptime commented 1 year ago

i thought the continue on error was supposed to just keep running tests on all platforms even when one fails (to make sure its not isolated to windows or something) but maybe im wrong

the good news is it looks like the issue is limited to that single inference test (based on individual-level permutation) so i'll take a closer look at that function today

knaaptime commented 1 year ago

(im fine to go ahead and merge this PR then and work on the failure separately)

jGaboardi commented 1 year ago

Looks like this started failing 5 days ago.

martinfleis commented 1 year ago

i thought the continue on error was supposed to just keep running tests on all platforms even when one fails

That is fail-fast.