[
https://issues.apache.org/jira/browse/SPARK-10385?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15096859#comment-15096859
]
Xiangrui Meng commented on SPARK-10385:
---------------------------------------
Spark 1.6 contains covariance and Pearson's correlation, but there are other
useful bivariate statistics.
> Bivariate statistics in DataFrames
> ----------------------------------
>
> Key: SPARK-10385
> URL: https://issues.apache.org/jira/browse/SPARK-10385
> Project: Spark
> Issue Type: Umbrella
> Components: ML, SQL
> Reporter: Xiangrui Meng
> Assignee: Burak Yavuz
>
> Similar to SPARK-10384, it would be nice to have bivariate statistics support
> in DataFrames (defined as UDAFs). This JIRA discuss general implementation
> and track subtasks. Bivariate statistics include:
> * continuous: covariance (SPARK-9297), Pearson's correlation (SPARK-9298),
> and Spearman's correlation (SPARK-10645)
> * categorical: ??
> If we define them as UDAFs, it would be flexible to use them with DataFrames,
> e.g.,
> {code}
> df.groupBy("key").agg(corr("x", "y"))
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
