Github user yanboliang commented on a diff in the pull request:
https://github.com/apache/spark/pull/8866#discussion_r48646093
--- Diff:
mllib/src/main/scala/org/apache/spark/ml/regression/DecisionTreeRegressor.scala
---
@@ -129,6 +136,22 @@ final class DecisionTreeRegressionModel private[ml] (
rootNode.predictImpl(features).prediction
}
+ /** We need to update this function if we ever add other impurity
measures. */
+ protected def predictVariance(features: Vector): Double = {
+ rootNode.predictImpl(features).impurityStats.calculate()
+ }
+
+ override protected def transformImpl(dataset: DataFrame): DataFrame = {
--- End diff --
Good points! I have updated my code and add test for turning off
```predictionCol``` but turning on ```varianceCol```, it works as expected.
I have noticed that it will output warning log for turning off
```predictionCol``` in
[```PredictionModel```](https://github.com/apache/spark/blob/master/mllib/src/main/scala/org/apache/spark/ml/Predictor.scala#L178),
I think we should remove it because if users set ```predictionCol``` to
```""```(null string) means they don't want to output ```predictionCol``` in
purpose and we should not notice them. However this issue is not relevant to
this PR, I will file another JIRA to track it.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
