[
https://issues.apache.org/jira/browse/SYSTEMML-1035?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15573393#comment-15573393
]
Glenn Weidner edited comment on SYSTEMML-1035 at 10/13/16 10:26 PM:
--------------------------------------------------------------------
The failures in DataFrameVectorScriptTest and
DataFrameVectorFrameConversionTest occur when calling df.first() at
FrameRDDConverterUtils.dataFrameToBinaryBlock(FrameRDDConverterUtils.java:352)
where df is DataFrame.
The failures in MLContextTest occur when calling tmp.count() at
RDDConverterUtils.java:290 where tmp is JavaRDD<Row>.
was (Author: gweidner):
The failures in DataFrameVectorScriptTest and
DataFrameVectorFrameConversionTest occur when calling df.first() at
FrameRDDConverterUtils.dataFrameToBinaryBlock(FrameRDDConverterUtils.java:352).
The failures in MLContextTest occur when calling tmp.count() at
RDDConverterUtils.java:290 where tmp is JavaRDD<Row>.
> ClassCastException's in unit tests
> ----------------------------------
>
> Key: SYSTEMML-1035
> URL: https://issues.apache.org/jira/browse/SYSTEMML-1035
> Project: SystemML
> Issue Type: Sub-task
> Components: Runtime, Test
> Reporter: Glenn Weidner
>
> Failures occurred in DataFrameVectorScriptTest,
> DataFrameVectorFrameConversionTest, MLContextFrameTest, and MLContextTest due
> to ClassCastException.
> Common example:
> java.lang.RuntimeException: java.lang.ClassCastException: java.lang.Integer
> cannot be cast to java.lang.Double
> originating in:
> org.apache.spark.sql.catalyst
> triggered from:
> org.apache.sysml.runtime.instructions.spark.utils.FrameRDDConverterUtils.dataFrameToBinaryBlock(FrameRDDConverterUtils.java:352)
> To reproduce, change pom.xml spark.version to 1.6.2 and run either of the
> four failing test suites.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
