[
https://issues.apache.org/jira/browse/SPARK-34037?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17260270#comment-17260270
]
Kent Yao edited comment on SPARK-34037 at 1/15/21, 5:43 PM:
------------------------------------------------------------
Oh, I mean that when users query for 4 attributes, for example, we may return 5
(or maybe more) instead, with one attribute named `aggOrder`.
This attribute is added by spark's analyzer for push *sortorder * into
aggeration for internal logic only and it should not affect the final output
*The above comment is outdated as we change the JIRA title and desc*
was (Author: qin yao):
Oh, I mean that when users query for 4 attributes, for example, we may return 5
(or maybe more) instead, with one attribute named `aggOrder`.
This attribute is added by spark's analyzer for push *sortorder * into
aggeration for internal logic only and it should not affect the final output
> Remove unnecessary upcasting for Avg & Sum which handle by themself internally
> ------------------------------------------------------------------------------
>
> Key: SPARK-34037
> URL: https://issues.apache.org/jira/browse/SPARK-34037
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 3.1.0
> Reporter: Kent Yao
> Priority: Major
>
> The type-coercion for numeric types of average and sum is not necessary at
> all, as the resultType and sumType can prevent the overflow.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
