Github user srowen commented on a diff in the pull request:
https://github.com/apache/spark/pull/20632#discussion_r169736202
--- Diff: mllib/src/main/scala/org/apache/spark/ml/tree/Node.scala ---
@@ -287,6 +291,34 @@ private[tree] class LearningNode(
}
}
+ /**
+ * @return true iff the node is a leaf.
+ */
+ private def isLeafNode(): Boolean = leftChild.isEmpty &&
rightChild.isEmpty
+
+ // the set of (leaf) predictions appearing in the subtree rooted at the
given node.
+ private lazy val leafPredictions: Set[Double] = {
--- End diff --
Imagine a full binary tree of 2 levels (7 nodes) with all identical
predictions. It can be pruned down to one node, the root. But I think this
logic would only prune the lowest level of nodes. Right? you're matching on
nodes that have `LeafNode` children only.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
