[
https://issues.apache.org/jira/browse/SPARK-26050?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16688955#comment-16688955
]
Hyukjin Kwon commented on SPARK-26050:
--------------------------------------
It's easily able to work around. Currently Spark has too many APIs open. Let's
avoid to add new APIs unless it's strongly needed.
> Implment withColumnExpr method on DataFrame
> -------------------------------------------
>
> Key: SPARK-26050
> URL: https://issues.apache.org/jira/browse/SPARK-26050
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 2.4.0
> Reporter: Mathew
> Priority: Major
>
> Currently we provide some syntactic sugar in the form of df.selectExpr(),
> which effectively executes as df.select(expr(), expr(), ...)
> I propose we implement a df.withColumnExpr(), which behaves similarly to
> df.withColumn(), except without the colName parameter, instead taking column
> names from the expressions themselves.
> This would stop the unfriendly paradigm of chained
> .withColumn().withColumn().withColumn() expressions, as we could allow
> passing as many column expressions as you want.
> Similar to df.selectExpr(), we should support all of: 'column names', 'column
> expressions', 'column string expressions' as inputs.
> Comments are welcome.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
