Github user sethah commented on a diff in the pull request:
https://github.com/apache/spark/pull/19638#discussion_r148662710
--- Diff:
mllib/src/main/scala/org/apache/spark/mllib/stat/MultivariateOnlineSummarizer.scala
---
@@ -230,6 +230,13 @@ class MultivariateOnlineSummarizer extends
MultivariateStatisticalSummary with S
override def count: Long = totalCnt
/**
+ * Number of parameters
+ *
+ */
+ @Since("2.3.0")
+ override def numParam: Int = n
--- End diff --
Hmmm... no this isn't right. When this is used with regression metrics, the
summary is of dimension 2 always since the `predictionsAndObservations` is just
an rdd of (prediction, label) pairs. The unit tests weren't thorough enough to
catch it.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
