[
https://issues.apache.org/jira/browse/TOREE-166?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15204828#comment-15204828
]
Gino Bustelo commented on TOREE-166:
------------------------------------
The code recommended is not enough to get this working. The issue is that
PySpark wants the sqlcontext and sparkContext to be an actual Python type that
handles crossing to the JVM.
Keep track of the progress of this work on PR
https://github.com/apache/incubator-toree/pull/15
> sqlContext not shared with PySpark and sparkR
> ---------------------------------------------
>
> Key: TOREE-166
> URL: https://issues.apache.org/jira/browse/TOREE-166
> Project: TOREE
> Issue Type: Bug
> Reporter: nimbusgo
> Assignee: Gino Bustelo
> Fix For: 0.1.0
>
>
> The scala interpreter and sql interpreter appear to share the same sqlContext
> and you can select tables in the sql interpreter that were registered in the
> scala interpreter. However, It appears that the PySpark and SparkR
> interpreters each create their own sqlContext on construction, and dataframes
> registered on those sqlContext will not be shared with the sqlContext in
> other interpreters. Would it be possible to change it so that the python and
> R interpreters were instantiated with the same sqlContext as the scala
> interpreter?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
