BUG: Offset-based rolling window, with only one raw in dataframe and closed='left', max and min functions make python crash #24811
Conversation
Codecov Report
@@ Coverage Diff @@
## master #24811 +/- ##
=======================================
Coverage 92.38% 92.38%
=======================================
Files 166 166
Lines 52382 52382
=======================================
Hits 48395 48395
Misses 3987 3987
Continue to review full report at Codecov.
|
Codecov Report
@@ Coverage Diff @@
## master #24811 +/- ##
==========================================
- Coverage 92.38% 92.38% -0.01%
==========================================
Files 166 166
Lines 52382 52379 -3
==========================================
- Hits 48395 48391 -4
- Misses 3987 3988 +1
Continue to review full report at Codecov.
|
|
Hi @jschendel, no worries. The entry has been added now. |
| @@ -1808,6 +1808,7 @@ Plotting | |||
| Groupby/Resample/Rolling | |||
| ^^^^^^^^^^^^^^^^^^^^^^^^ | |||
|
|
|||
| - Bug in :func:`pandas.core.window.Rolling.min` and :func:`pandas.core.window.Rolling.max` with ``closed='left'``, a datetime-like index and only one entry in the series leading to segfault (:issue:`24718`) | |||
There was a problem hiding this comment.
Is this truly this specific? Off the top of my head seems strange that this would only be applicable to date time like indices
There was a problem hiding this comment.
The datetime-like index makes the program going into the function _roll_min_max_variable of window.pyx, where the bug is. I am not very sure if there is any other case triggering the function. But the program normally goes into _roll_min_max_fixed if it isn't a datetime-like index.
jreback
changed the title
|
thanks @jh-wu |
…closed='left', max and min functions make python crash (pandas-dev#24811)
…closed='left', max and min functions make python crash (pandas-dev#24811)
git diff upstream/master -u -- "*.py" | flake8 --diff