Github user vlad17 commented on a diff in the pull request:
https://github.com/apache/spark/pull/14547#discussion_r78484674
--- Diff: mllib/src/main/scala/org/apache/spark/ml/tree/treeParams.scala ---
@@ -183,24 +191,18 @@ private[ml] trait DecisionTreeParams extends
PredictorParams
*/
private[ml] trait TreeClassifierParams extends Params {
+ // Impurity should be overriden when setting a default. This should be a
def, but has
+ // to be a val to maintain the proper documentation.
/**
- * Criterion used for information gain calculation (case-insensitive).
- * Supported: "entropy" and "gini".
- * (default = gini)
* @group param
*/
- final val impurity: Param[String] = new Param[String](this, "impurity",
"Criterion used for" +
- " information gain calculation (case-insensitive). Supported options:"
+
- s" ${TreeClassifierParams.supportedImpurities.mkString(", ")}",
- (value: String) =>
TreeClassifierParams.supportedImpurities.contains(value.toLowerCase))
-
- setDefault(impurity -> "gini")
+ val impurity: Param[String] = new Param(this, "", "")
--- End diff --
right, that's the point... we have loss-based default for gbts, in contrast
to other trees
---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
