[ https://issues.apache.org/jira/browse/SPARK-5752?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14317761#comment-14317761 ]
Apache Spark commented on SPARK-5752: ------------------------------------- User 'rxin' has created a pull request for this issue: https://github.com/apache/spark/pull/4556 > Don't implicitly convert RDDs directly to DataFrames > ---------------------------------------------------- > > Key: SPARK-5752 > URL: https://issues.apache.org/jira/browse/SPARK-5752 > Project: Spark > Issue Type: Sub-task > Components: SQL > Reporter: Reynold Xin > Assignee: Reynold Xin > > DataFrame is a rich API consisting of too many functions. It would be safer > to convert RDDs to a DataFrameHolder that consists of only two functions: > {code} > - toDataFrame() > - toDataFrame(String*) > {code} > This way, it is highly unlikely to have ambiguous implicit conversions, at > the cost of requiring users to always call toDataFrame before being able to > use DataFrame functions. -- 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