[
https://issues.apache.org/jira/browse/FLINK-6242?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15973419#comment-15973419
]
ASF GitHub Bot commented on FLINK-6242:
---------------------------------------
Github user fhueske commented on a diff in the pull request:
https://github.com/apache/flink/pull/3735#discussion_r111994200
--- Diff:
flink-libraries/flink-table/src/main/scala/org/apache/flink/table/runtime/aggregate/DataSetPreAggFunction.scala
---
@@ -64,38 +68,22 @@ class DataSetPreAggFunction(
def preaggregate(records: Iterable[Row], out: Collector[Row]): Unit = {
--- End diff --
I think we can move the implementation of `preaggregate()` to `combine()`
and let `mapPartition()` call `combine()`. `combine()` is called for groups for
records with the key and `mapPartition()` just once (for the whole partition).
This way we can remove some overhead from `combine()`.
> codeGen DataSet Goupingwindow Aggregates
> ----------------------------------------
>
> Key: FLINK-6242
> URL: https://issues.apache.org/jira/browse/FLINK-6242
> Project: Flink
> Issue Type: Sub-task
> Components: Table API & SQL
> Reporter: Shaoxuan Wang
> Assignee: Shaoxuan Wang
>
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
