GitHub user WeichenXu123 opened a pull request:
https://github.com/apache/spark/pull/20786
[SPARK-14681][ML] Provide label/impurity stats for spark.ml decision tree
nodes
## What changes were proposed in this pull request?
API:
```
trait ClassificationNode extends Node
def getLabelCount(label: Int): Double
trait RegressionNode extends Node
def getCount(): Double
def getSum(): Double
def getSquareSum(): Double
// turn LeafNode to be trait
trait LeafNode extends Node {
def prediction: Double
def impurity: Double
...
}
class ClassificationLeafNode extends ClassificationNode with LeafNode
class RegressionLeafNode extends RegressionNode with LeafNode
// turn InternalNode to be trait
trait InternalNode extends Node{
def gain: Double
def leftChild: Node
def rightChild: Node
def split: Split
...
}
class ClassificationInternalNode extends ClassificationNode with
InternalNode
class RegressionInternalNode extends RegressionNode with InternalNode
```
### Note: this class hierarchy change will break binary compatibility but
will keep source compatibility.
## How was this patch tested?
UT will be added soon.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/WeichenXu123/spark tree_stat_api_2
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/spark/pull/20786.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #20786
----
commit 9dacad30df747add2317d10484550f8f0173a672
Author: WeichenXu <weichen.xu@...>
Date: 2018-03-09T14:38:30Z
init pr
----
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
