[
https://issues.apache.org/jira/browse/SPARK-50846?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17916195#comment-17916195
]
Pengfei Xu edited comment on SPARK-50846 at 1/22/25 9:37 PM:
-------------------------------------------------------------
Hey [~ostronaut], thanks for helping me! But this task depends on a PR that is
not yet merged:
https://github.com/apache/spark/pull/49111/files#diff-bb66ac05139568572b3fe701a1e44087ab6de77095800b96a009e365e408d380R175-R184.
, so I suggest to wait for a while.
was (Author: JIRAUSER304650):
Hey [~ostronaut], this task depends on a PR that is not yet merged:
https://github.com/apache/spark/pull/49111/files#diff-bb66ac05139568572b3fe701a1e44087ab6de77095800b96a009e365e408d380R175-R184.
I suggest to wait for it.
> Consolidate Aggregator handling in ColumnNodeToProtoConverter
> -------------------------------------------------------------
>
> Key: SPARK-50846
> URL: https://issues.apache.org/jira/browse/SPARK-50846
> Project: Spark
> Issue Type: Improvement
> Components: Connect
> Affects Versions: 4.0.0
> Reporter: Pengfei Xu
> Priority: Major
>
> We now have two matches for
> {code:java}
> InvokeInlineUserDefinedFunction(aggregator, args, ...)
> {code}
> in ColumnNodeToProtoConverter, one with args being empty and one not.
> The empty case is for`df.select(agg)` and the second case is for
> `kvds.agg(agg)`. We should merge them into one.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
