[
https://issues.apache.org/jira/browse/SPARK-33303?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Hyukjin Kwon resolved SPARK-33303.
----------------------------------
Fix Version/s: 3.1.0
Resolution: Fixed
Issue resolved by pull request 30203
[https://github.com/apache/spark/pull/30203]
> Deduplicate deterministic PythonUDF calls
> -----------------------------------------
>
> Key: SPARK-33303
> URL: https://issues.apache.org/jira/browse/SPARK-33303
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 3.1.0
> Reporter: Peter Toth
> Assignee: Apache Spark
> Priority: Major
> Fix For: 3.1.0
>
>
> We run into an issue where a customer created a column with an expensive
> PythonUDF call and build a very complex logic on the the top of that column
> as new derived columns. Due to `CollapseProject` and `ExtractPythonUDFs`
> rules the UDF is called ~1000 times for each row which degraded the
> performance of the query significantly.
> The `ExtractPythonUDFs` rule could deduplicate deterministic UDFs so as to
> avoid performance degradation.
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
