[
https://issues.apache.org/jira/browse/SPARK-17781?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15566457#comment-15566457
]
Shivaram Venkataraman commented on SPARK-17781:
-----------------------------------------------
[~falaki] I looked at this a bit more today and it looks like this problem is
specific to dapplyCollect -- what happens here is that we dont have the schema
for the output table, so we serialize it as a byteArray [1] and rely on the
driver to do the conversion / deserialization while running collect. I couldn't
trace this part to the end, but it looks like this gets deserialized in [2] and
the call to unserialize there interprets the bytes as double instead of date.
I'm not sure what is a good fix for this as well.
[1] https://github.com/apache/spark/blob/master/R/pkg/inst/worker/worker.R#L75
[2] https://github.com/apache/spark/blob/master/R/pkg/R/DataFrame.R#L1431
> datetime is serialized as double inside dapply()
> ------------------------------------------------
>
> Key: SPARK-17781
> URL: https://issues.apache.org/jira/browse/SPARK-17781
> Project: Spark
> Issue Type: Bug
> Components: SparkR
> Affects Versions: 2.0.0
> Reporter: Hossein Falaki
>
> When we ship a SparkDataFrame to workers for dapply family functions, inside
> the worker DateTime objects are serialized as double.
> To reproduce:
> {code}
> df <- createDataFrame(data.frame(id = 1:10, date = Sys.Date()))
> dapplyCollect(df, function(x) { return(x$date) })
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
