You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Bryan Cutler (JIRA)" <ji...@apache.org> on 2017/09/22 16:57:00 UTC
[jira] [Commented] (SPARK-22106) Remove support for 0-parameter
pandas_udfs
[ https://issues.apache.org/jira/browse/SPARK-22106?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16176714#comment-16176714 ]
Bryan Cutler commented on SPARK-22106:
--------------------------------------
I'll submit a PR soon for this
> Remove support for 0-parameter pandas_udfs
> ------------------------------------------
>
> Key: SPARK-22106
> URL: https://issues.apache.org/jira/browse/SPARK-22106
> Project: Spark
> Issue Type: Improvement
> Components: PySpark, SQL
> Affects Versions: 2.3.0
> Reporter: Bryan Cutler
>
> SPARK-21190 Added support for vectorized pandas_udfs and allows a 0-parameter udf to be created, but it must accept **kwargs to get the correct length of the vector to return. This behavior is a confusing and awkward to implement and the overall pandas_udf usage would be much cleaner if this was not allowed.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org