Open johannjacobsohn opened 5 years ago
Thanks @johannjacobsohn for pointing out and fixing the pandas deprecation issue that caused pyplis to fail (fixed in latest release. Regarding version pinning: I am not sure... I would like to avoid strict pinning (i.e. equality) but I am happy to specify >= or similar. Maybe the better strategy would be to setup travis such, that it runs the build and tests in specified time intervals, or similar. Don't know if that's possible, but I can check that. This would ofc benefit from better test coverage (cf. #29) ;)
What do you think?
Pandas 0.25 broke pyplis build because it removed from_csv. Sudden outside breakage is pretty annoying and is usually avoided by specifying supported dependency versions, eg in
pyplis_env.yml
. What do you think?