You are viewing a plain text version of this content. The canonical link for it is here.
Posted to jira@arrow.apache.org by "Ben Kietzman (Jira)" <ji...@apache.org> on 2021/08/30 17:44:00 UTC
[jira] [Resolved] (ARROW-12725) [C++][Compute] GroupBy: improve
performance by encoding keys in row format only when they are inserted into
hash table
[ https://issues.apache.org/jira/browse/ARROW-12725?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Ben Kietzman resolved ARROW-12725.
----------------------------------
Resolution: Fixed
Issue resolved by pull request 10290
[https://github.com/apache/arrow/pull/10290]
> [C++][Compute] GroupBy: improve performance by encoding keys in row format only when they are inserted into hash table
> ----------------------------------------------------------------------------------------------------------------------
>
> Key: ARROW-12725
> URL: https://issues.apache.org/jira/browse/ARROW-12725
> Project: Apache Arrow
> Issue Type: Improvement
> Components: C++
> Affects Versions: 4.0.0
> Reporter: Michal Nowakiewicz
> Assignee: Michal Nowakiewicz
> Priority: Major
> Labels: pull-request-available
> Fix For: 6.0.0
>
> Time Spent: 1h 40m
> Remaining Estimate: 0h
>
> Previous implementation of hash group by converts input ExecBatches to row-oriented format,
> then hashes and compares rows as if they were a single column.
> It is more efficient (especially for small number of key columns) to avoid relatively costly
> encoding and instead compute hashes of individual columns in column-oriented format mixing them together, and similarly comparing column-oriented data to row-oriented data in the hash table without converting.
> Encoding only happens for a subset of input rows that are inserted into the hash table - they introduce new groups.
> Keys in hash table remain stored as row-oriented.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)