-
Notifications
You must be signed in to change notification settings - Fork 2
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
CI tests fail when upgrading pandas from 2.1.4 to >=2.2.0 #232
Comments
Looks like all the
|
Looks like it's started to fail since it tried to upgrade pandas to v2.2.0. I updated the title and description with this info. |
I think it's probably related to this: https://pandas.pydata.org/docs/whatsnew/v2.2.0.html#merge-and-dataframe-join-now-consistently-follow-documented-sort-behavior |
#216 now includes a fix. |
I appreciate the fix @JMGilbert! Stop me if I'm reading this wrong but on quick glance it looks like the issue isn't with the dscim package code itself. The issue is more that the tests were written to read the expected results from an older file of stashed results. This expected results file was written with an older version of pandas and so it needs to get sorted before it can be compared with the results generated from the newer pandas version? Looking at this ... hmm... It's not something that needs to get fixed here, but It would be ideal if the tests didn't need to handle this type of logic, instead focusing on what's needed to test some target behavior. The hacky workaround stuff makes it harder to reason what the test is actually trying to test. It's going to keep getting muddled as time marches forward. Again, I think this is absolutely something that shouldn't be addressed in this issue and PR. The code that needs updating had this hacky stuff before this issue. Depending on priorities, this is maybe something to consider for future cleanup before legacy slows things down too much, @kemccusker. It might come back to bite later. |
@brews would love your input on how better to write the tests to avoid these types of issues, for future cleanup. For this PR, sorting seems like a reasonable fix, yeah? |
Closed by #216 |
PR #216 suggests the package fails tests when the CI environment bumps pandas from v2.1.4 to the latest version (currently v2.2.1). It started failing since the pandas v2.2.0 release.
I can't judge why it's failing from a quick look at the testing logs.
Archived logs from the failed CI run are here: logs_22096399592.zip
The text was updated successfully, but these errors were encountered: