[ https://issues.apache.org/jira/browse/SPARK-21416?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16087161#comment-16087161 ]
Nick Pentreath edited comment on SPARK-21416 at 7/14/17 10:37 AM: ------------------------------------------------------------------ This is in Spark 2.2, which also adds probability. was (Author: mlnick): This is in Spark 2.2 > Emit Raw Prediction Scores in DecisionTree Classification Model > --------------------------------------------------------------- > > Key: SPARK-21416 > URL: https://issues.apache.org/jira/browse/SPARK-21416 > Project: Spark > Issue Type: Improvement > Components: ML > Affects Versions: 2.0.1 > Reporter: Aman Rawat > > DecisionTreeClassificationModel today emits only the final prediction score > (index for the class). However, it's often required to know the raw > prediction score for each class as well. This is already available in the > underlying model, but not exposed in "ML" (i.e. DataFrame based) interace - > DecisionTreeClassificationModel. > The transformer's output can be changed to include both "prediction" and > "rawPrediction" columns. -- This message was sent by Atlassian JIRA (v6.4.14#64029) --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org