[
https://issues.apache.org/jira/browse/SPARK-58637?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Ruifeng Zheng resolved SPARK-58637.
-----------------------------------
Fix Version/s: 4.4.0
Resolution: Fixed
Issue resolved by pull request 57845
[https://github.com/apache/spark/pull/57845]
> Optimize HashingTF transform closure
> ------------------------------------
>
> Key: SPARK-58637
> URL: https://issues.apache.org/jira/browse/SPARK-58637
> Project: Spark
> Issue Type: Sub-task
> Components: ML
> Affects Versions: 5.0.0
> Reporter: Ruifeng Zheng
> Assignee: Ruifeng Zheng
> Priority: Major
> Labels: pull-request-available
> Fix For: 4.4.0
>
>
> HashingTF.transform calls the instance indexOf method from its UDF, which
> captures the transformer and looks up numFeatures for every term. Select the
> hash and binary update functions before constructing the UDF, calculate
> indexes from captured local values, store term counts as Int, and convert
> them to Double only when constructing the output vector.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
