You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hive.apache.org by "Jesus Camacho Rodriguez (Jira)" <ji...@apache.org> on 2020/10/06 03:31:00 UTC
[jira] [Commented] (HIVE-24232) Incorrect translation of rollup
expression from Calcite
[ https://issues.apache.org/jira/browse/HIVE-24232?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17208469#comment-17208469 ]
Jesus Camacho Rodriguez commented on HIVE-24232:
------------------------------------------------
The PR also adds printing of grouping sets for the Hive Group By operators in the Hive plan.
> Incorrect translation of rollup expression from Calcite
> -------------------------------------------------------
>
> Key: HIVE-24232
> URL: https://issues.apache.org/jira/browse/HIVE-24232
> Project: Hive
> Issue Type: Bug
> Components: CBO
> Reporter: Jesus Camacho Rodriguez
> Assignee: Jesus Camacho Rodriguez
> Priority: Major
>
> In Calcite, it is not necessary that the columns in the group set are in the same order as the rollup. For instance, this is the Calcite representation of a rollup for a given query:
> {code}
> HiveAggregate(group=[{1, 6, 7}], groups=[[{1, 6, 7}, {1, 7}, {1}, {}]], agg#0=[sum($12)], agg#1=[count($12)], agg#2=[sum($4)], agg#3=[count($4)], agg#4=[sum($15)], agg#5=[count($15)])
> {code}
> When we generate the Hive plan from the Calcite operator, we make such assumption incorrectly.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)