You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Xinrong Meng (Jira)" <ji...@apache.org> on 2022/12/08 23:59:00 UTC
[jira] [Updated] (SPARK-41455) Resolve dtypes inconsistencies of date/timestamp functions
[ https://issues.apache.org/jira/browse/SPARK-41455?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Xinrong Meng updated SPARK-41455:
---------------------------------
Description:
When implementing date/timestamp functions, we notice inconsistent dtypes with PySpark, as shown below.
{code:python}
>> sdf.select(SF.current_timestamp()).toPandas().dtypes
current_timestamp() datetime64[ns]
dtype: object
>>> cdf.select(CF.current_timestamp()).toPandas().dtypes
current_timestamp() datetime64[ns, America/Los_Angeles]
{code}
Affected functions include: `to_timestamp, from_utc_timestamp, to_utc_timestamp, timestamp_seconds, current_timestamp`.
We may have to implement `is_timestamp_ntz_preferred` for Connect.
After the fix, tests of those date/timestamp functions which use `compare_by_show` should be switched to `toPandas` comparison.
was:
When implementing date/timestamp functions, we notice inconsistent dtypes with PySpark, as shown below.
{code:python}
>> sdf.select(SF.current_timestamp()).toPandas().dtypes
current_timestamp() datetime64[ns]
dtype: object
>>> cdf.select(CF.current_timestamp()).toPandas().dtypes
current_timestamp() datetime64[ns, America/Los_Angeles]
{code}
We may have to implement `is_timestamp_ntz_preferred` for Connect.
After the fix, tests of those date/timestamp functions which use `compare_by_show` should be switched to `toPandas` comparison.
> Resolve dtypes inconsistencies of date/timestamp functions
> ----------------------------------------------------------
>
> Key: SPARK-41455
> URL: https://issues.apache.org/jira/browse/SPARK-41455
> Project: Spark
> Issue Type: Sub-task
> Components: PySpark
> Affects Versions: 3.4.0
> Reporter: Xinrong Meng
> Priority: Major
>
> When implementing date/timestamp functions, we notice inconsistent dtypes with PySpark, as shown below.
> {code:python}
> >> sdf.select(SF.current_timestamp()).toPandas().dtypes
> current_timestamp() datetime64[ns]
> dtype: object
> >>> cdf.select(CF.current_timestamp()).toPandas().dtypes
> current_timestamp() datetime64[ns, America/Los_Angeles]
> {code}
> Affected functions include: `to_timestamp, from_utc_timestamp, to_utc_timestamp, timestamp_seconds, current_timestamp`.
> We may have to implement `is_timestamp_ntz_preferred` for Connect.
> After the fix, tests of those date/timestamp functions which use `compare_by_show` should be switched to `toPandas` comparison.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org