[REVIEW] Reindex in DataFrame.__setitem__ - #7948
Conversation
|
Does this change fix the following build error perchance? |
Yes, CI is currently blocked on this test failure and this PR addresses it. |
Codecov Report
@@ Coverage Diff @@
## branch-0.20 #7948 +/- ##
===============================================
+ Coverage 82.30% 82.72% +0.42%
===============================================
Files 101 103 +2
Lines 17053 17714 +661
===============================================
+ Hits 14035 14654 +619
- Misses 3018 3060 +42
Continue to review full report at Codecov.
|
|
@gpucibot merge |
This PR fixes missing reindexing in `DataFrame.__setitem__` when the `value` argument is a `DataFrame`, we currently align index if `value` is a Series & `arg` is a column name already. This change is necessary to continue with the upgrade to pandas `1.2.4`, however pandas has confirmed this as a regression only in `1.2.0` and `1.2.2` hence corrected the pytest to only xfail in those versions of pandas. Authors: - GALI PREM SAGAR (https://github.com/galipremsagar) Approvers: - Ashwin Srinath (https://github.com/shwina) URL: NVIDIA#7948
* Reindex in `DataFrame.__setitem__` (#7948) This PR fixes missing reindexing in `DataFrame.__setitem__` when the `value` argument is a `DataFrame`, we currently align index if `value` is a Series & `arg` is a column name already. This change is necessary to continue with the upgrade to pandas `1.2.4`, however pandas has confirmed this as a regression only in `1.2.0` and `1.2.2` hence corrected the pytest to only xfail in those versions of pandas. Authors: - GALI PREM SAGAR (https://github.com/galipremsagar) Approvers: - Ashwin Srinath (https://github.com/shwina) URL: #7948 * pin max pandas to 1.2.4
This PR fixes missing reindexing in
DataFrame.__setitem__when thevalueargument is aDataFrame, we currently align index ifvalueis a Series &argis a column name already.This change is necessary to continue with the upgrade to pandas
1.2.4, however pandas has confirmed this as a regression only in1.2.0and1.2.2hence corrected the pytest to only xfail in those versions of pandas.