[
https://issues.apache.org/jira/browse/SPARK-36263?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17676265#comment-17676265
]
Nick Hryhoriev commented on SPARK-36263:
----------------------------------------
Hi, I find this feature does not work for `foreach` or `foreachPartition`
actions.
Maybe because they use `rdd.foreach` under the hood.
Look like it's because `foreach` or `foreachPartition` actions do not work with
`QueryExecutionListener`.
Example to reproduce
[https://gist.github.com/GrigorievNick/e7cf9ec5584b417d9719e2812722e6d3]
Do I miss something or is it's a know issue?
> Add Dataset.observe(Observation, Column, Column*) to PySpark
> ------------------------------------------------------------
>
> Key: SPARK-36263
> URL: https://issues.apache.org/jira/browse/SPARK-36263
> Project: Spark
> Issue Type: New Feature
> Components: PySpark
> Affects Versions: 3.3.0
> Reporter: Enrico Minack
> Assignee: Enrico Minack
> Priority: Major
> Fix For: 3.3.0
>
>
> With SPARK-34806 we now have a way to use the `Dataset.observe` method
> without the need to interact with
> `org.apache.spark.sql.util.QueryExecutionListener`. This allows us to easily
> retrieve observations in PySpark.
> Adding a `Dataset.observe(Observation, Column, Column*)` equivalent to
> PySpark's `DataFrame` is straightforward while it allows to utilise
> observations from Python.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
