BUG: downcast = 'unsigend' on 0 would would not downcast to unsigned. #14472
Conversation
|
@verhalenn For future reference, it is normally never needed to open a new PR to fix git issues. You can just clean it up locally and force push. BTW, you have some failing tests. |
|
No when I forced pushed it it closed the pull request for some reason. Usually that would happen if I were to delete the branch and then force push but that wasn't the case. I'll look into the error. |
Current coverage is 85.25% (diff: 100%)
|
| @@ -401,6 +401,32 @@ def test_downcast(self): | |||
| res = pd.to_numeric(data, downcast=downcast) | |||
| tm.assert_numpy_array_equal(res, expected) | |||
|
|
|||
| # check that the smallest and largest values in each integer type pass to each type. | |||
There was a problem hiding this comment.
can you make this into a separate test function
add the issue number as a comment
ping on green
| - Bug in ``pd.pivot_table`` may raise ``TypeError`` or ``ValueError`` when ``index`` or ``columns`` | ||
| is not scalar and ``values`` is not specified (:issue:`14380`) |
There was a problem hiding this comment.
you eliminated this line, pls restore it (its another issue)
I also added a series of other tests that included making sure that a dtype shifted dtype once it reached the next value and once again edited the whatsnew.
This reverts commit 636de09. Conflicts: doc/source/whatsnew/v0.19.1.txt
|
thanks! |
|
@jreback : I think this commit is breaking |
|
reverted: d1d75d7 @verhalenn if you can re-submit this PR would be great. |
|
First thing tomorrow morning. I was kind of surprised you closed it when you did. The lint should be an easy fix but the downcast test will take a little extra time. |
Continuation of #14412
git diff upstream/master | flake8 --diffOnce again I had some very bizarre issues with git. Hope this does it this time.