Jon Niehof
Jon Niehof
Also since this usedtofix #484 now, need to update the PR description to note that. EDIT: No longer true, removing magic word.
It looks like @balarsen pushed some commits to branch balarsen-toolbox_doc instead of toolbox_doc. I have: - pulled those commits to this toolbox_doc - rebased the toolbox_doc branch against its original...
Unless someone objects, this will be merged no earlier than Friday 23 September
Does 1.2 support what we do? Then we could wrap and thunk down to 1.2 as a prelude to deprecation and removal (at the point where we're ready to require...
It would be good to see the other examples @balarsen has been using where scalar input results in 1D array output, since that's pretty uncommon in my experience. I'd expect...
Couple of possible approaches that come to mind: We could wrap the call to `unittest.main` in `test_all.py` with a `catch_warnings` to catch uncaught warnings and then fail. But I don't...
I think it would be possible for each test job to parse the output and look for warnings, then pass the information out to the "all tests" job which could...
From the upshot of #425, whatever course we take, we shouldn't do this on old versions of dependencies. Every Python version has an "oldest" and a "newest" dependency strategy in...
The return value is always in GEO. This is _not_ adequately documented. This has come up before and I thought it had been addressed, but apparently not. We've got a...
Or .zip, for that matter. loadpickle already does this sort of thing, so adapting that wouldn't be too bad.