Github user GayathriMurali commented on a diff in the pull request:
https://github.com/apache/spark/pull/12118#discussion_r58287249
--- Diff: mllib/src/main/scala/org/apache/spark/ml/tree/treeModels.scala ---
@@ -358,3 +376,100 @@ private[ml] object DecisionTreeModelReadWrite {
finalNodes.head
}
}
+
+private[ml] object EnsembleModelReadWrite {
+
+ /**
+ * Helper method for saving a tree ensemble to disk.
+ *
+ * @param instance Tree ensemble model
+ * @param path Path to which to save the ensemble model.
+ * @param extraMetadata Metadata such as numFeatures, numClasses,
numTrees.
+ */
+ def saveImpl[M <: Params with TreeEnsembleModel](
+ instance: M,
+ path: String,
+ sql: SQLContext,
+ extraMetadata: JObject): Unit = {
+ DefaultParamsWriter.saveMetadata(instance, path, sql.sparkContext,
Some(extraMetadata))
+ val treesMetadataJson: Array[(Int, String)] =
instance.trees.zipWithIndex.map {
+ case (tree, treeID) =>
+ treeID ->
DefaultParamsWriter.getMetadataToSave(tree.asInstanceOf[Params],
sql.sparkContext)
+ }
+ val treesMetadataPath = new Path(path, "treesMetadata").toString
+ sql.createDataFrame(treesMetadataJson).toDF("treeID", "metadata")
+ .write.parquet(treesMetadataPath)
+ val dataPath = new Path(path, "data").toString
+ val nodeDataRDD =
sql.sparkContext.parallelize(instance.trees.zipWithIndex).flatMap {
--- End diff --
Is it alright to use flatMap to combine RDDs? Can we use sparkContext.union
instead?
---
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]
