[
https://issues.apache.org/jira/browse/ARROW-13662?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17401105#comment-17401105
]
Joris Van den Bossche commented on ARROW-13662:
-----------------------------------------------
The bug in pandas was a difference in extracting the year from Index vs Series:
{code}
In [1]:
pd.to_datetime(["1970-01-01T00:00:59.123456789"]).tz_localize("UTC").tz_convert("US/Central").year
Out[1]: Int64Index([1969], dtype='int64')
In [2]:
pd.to_datetime(["1970-01-01T00:00:59.123456789"]).tz_localize("UTC").tz_convert("US/Central").to_series().dt.year
Out[2]:
1969-12-31 18:00:59.123456789-06:00 1970
dtype: int64
{code}
Fixed in pandas 1.0+ (we are using pandas as "reference" to compare the result
with)
> [CI] Failing test test_extract_datetime_components with pandas 0.24
> -------------------------------------------------------------------
>
> Key: ARROW-13662
> URL: https://issues.apache.org/jira/browse/ARROW-13662
> Project: Apache Arrow
> Issue Type: Bug
> Components: Continuous Integration, Python
> Reporter: Joris Van den Bossche
> Assignee: Joris Van den Bossche
> Priority: Major
> Labels: pull-request-available
> Time Spent: 10m
> Remaining Estimate: 0h
>
> See eg https://github.com/ursacomputing/crossbow/runs/3358009767
> Introduced by ARROW-12980, test is failing with pandas 0.24 due to a pandas
> bug.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
