[
https://issues.apache.org/jira/browse/SPARK-48478?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17850865#comment-17850865
]
Ian Cook commented on SPARK-48478:
----------------------------------
For Connect, see class {{LocalRelation}} in
{{{}python/pyspark/sql/connect/plan.py{}}}. Something similar could be used to
create a local relation from an iterator of RecordBatches. (But do we need to
create this as a cached remote relation? Creating it locally will just fill up
client memory I think)
For Classic, see {{{}_create_from_arrow_table{}}}. Something similar could be
used to create a DataFrame from an iterator of RecordBatches.
> Allow passing iterator of PyArrow RecordBatches to createDataFrame()
> --------------------------------------------------------------------
>
> Key: SPARK-48478
> URL: https://issues.apache.org/jira/browse/SPARK-48478
> Project: Spark
> Issue Type: Improvement
> Components: Connect, Input/Output, PySpark, SQL
> Affects Versions: 4.0.0, 3.5.1
> Reporter: Ian Cook
> Priority: Major
>
> As a follow-up to SPARK-48220:
> For larger data, it would be nice to be able to pass an iterator of PyArrow
> RecordBatches to {{{}createDataFrame(){}}}.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
