BUG: reshape fix for maybe_infer_to_datetimelike() #16395
Conversation
|
tests should always be included. generally I write them first to ensure that they are picking up the bug. you can push to this PR itself. |
jreback
added
Bug
Reshaping
Codecov Report
@@ Coverage Diff @@
## master #16395 +/- ##
==========================================
- Coverage 90.41% 90.4% -0.01%
==========================================
Files 161 161
Lines 50997 50997
==========================================
- Hits 46107 46106 -1
- Misses 4890 4891 +1
Continue to review full report at Codecov.
|
Codecov Report
@@ Coverage Diff @@
## master #16395 +/- ##
=======================================
Coverage 90.42% 90.42%
=======================================
Files 161 161
Lines 51033 51033
=======================================
Hits 46148 46148
Misses 4885 4885
Continue to review full report at Codecov.
|
|
@jreback I might need some guidance...I might not have needed to touch this -- the branch had a new conflict, I imagine because of other merged PRs, but only with whatsnew. Now I'm not sure how to fix it back to a state where tests pass. |
pandas/tests/dtypes/test_cast.py
Outdated
| @@ -213,6 +213,13 @@ def test_maybe_convert_scalar(self): | |||
| result = maybe_convert_scalar(Timedelta('1 day 1 min')) | |||
| assert result == Timedelta('1 day 1 min').value | |||
|
|
|||
| def test_maybe_infer_to_datetimelike(self): | |||
|
|
|||
pandas/tests/dtypes/test_cast.py
Outdated
| def test_maybe_infer_to_datetimelike(self): | ||
|
|
||
| # this construction raises IndexError in pandas=0.20.1 | ||
| result = DataFrame(np.array([[NaT, 'a', 'b', 0], |
There was a problem hiding this comment.
use both examples from here: #16362 (comment)
and assert the result to an expected result == 8
|
pls add a whatsnew note in 0.20.2 (bug fixes, reshaping section) |
|
Does cd11b49 look like the full set of changes, aside from the whatsnew? |
|
It looks like you lost some changes, but you should be able to recover them with your git reflog, assuming you haven't cleared it manually. Or maybe even a |
|
@TomAugspurger thanks, didn't have much time today and thankfully this is such a small change so I just redid the changes. I did add one more test assertion this morning upon Jeff's recommendation, and that's included here as well. |
|
thanks @RobinFiveWords |
closes pandas-dev#16362 Author: RobinFiveWords <[email protected]> Closes pandas-dev#16395 from RobinFiveWords/cast-infer-datetime-reshape-fix and squashes the following commits: 7ad1e7d [RobinFiveWords] redid lost changes to cast.py and test_cast.py afa2eeb [RobinFiveWords] added whatsnew0.20.2 entry 7a35624 [RobinFiveWords] removed whatsnew entry again 2ec60a6 [RobinFiveWords] added back whatsnew change (cherry picked from commit 05d0667)
closes #16362 Author: RobinFiveWords <[email protected]> Closes #16395 from RobinFiveWords/cast-infer-datetime-reshape-fix and squashes the following commits: 7ad1e7d [RobinFiveWords] redid lost changes to cast.py and test_cast.py afa2eeb [RobinFiveWords] added whatsnew0.20.2 entry 7a35624 [RobinFiveWords] removed whatsnew entry again 2ec60a6 [RobinFiveWords] added back whatsnew change (cherry picked from commit 05d0667)
closes pandas-dev#16362 Author: RobinFiveWords <[email protected]> Closes pandas-dev#16395 from RobinFiveWords/cast-infer-datetime-reshape-fix and squashes the following commits: 7ad1e7d [RobinFiveWords] redid lost changes to cast.py and test_cast.py afa2eeb [RobinFiveWords] added whatsnew0.20.2 entry 7a35624 [RobinFiveWords] removed whatsnew entry again 2ec60a6 [RobinFiveWords] added back whatsnew change
git diff upstream/master --name-only -- '*.py' | flake8 --diff