Barry Becker created SPARK-23824:
------------------------------------
Summary: Make inpurityStats publicly accessible in ml.tree.Node
Key: SPARK-23824
URL: https://issues.apache.org/jira/browse/SPARK-23824
Project: Spark
Issue Type: Wish
Components: ML
Affects Versions: 2.1.1
Reporter: Barry Becker
This is minor, but it is also a very easy fix.
I would like to visualize the structure of a decision tree model, but currently
the only means of obtaining the label distribution data at each node of the
tree is hidden within each ml.tree.Node inside the impurityStats.
I'm pretty sure that the fix for this is as easy as removing the private[ml]
qualifier from occurrences of
private[ml] def impurityStats: ImpurityCalculator
and
override private[ml] val impurityStats: ImpurityCalculator
As a workaround, I've put my class that needs access into a
org.apache.spark.ml.tree package in my own repository, but I would really like
to not have to do that.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
