Github user neggert commented on a diff in the pull request:
https://github.com/apache/spark/pull/14140#discussion_r70818220
--- Diff:
mllib/src/main/scala/org/apache/spark/mllib/regression/IsotonicRegression.scala
---
@@ -408,9 +409,11 @@ class IsotonicRegression private (private var
isotonic: Boolean) extends Seriali
*/
private def parallelPoolAdjacentViolators(
input: RDD[(Double, Double, Double)]): Array[(Double, Double,
Double)] = {
- val parallelStepResult = input
- .sortBy(x => (x._2, x._1))
- .glom()
+ val keyedInput = input.keyBy(_._2)
+ val parallelStepResult = keyedInput
+ .partitionBy(new RangePartitioner(keyedInput.getNumPartitions,
keyedInput))
+ .values
+ .mapPartitions( p => Iterator(p.toSeq.sortBy(x => (x._2,
x._1)).toArray))
--- End diff --
Actually, the `Iterator` is necessary. Since we removed the `glom`, we need
to return an `RDD[Array[T]]`. So the function passed to `mapPartitions` needs
to be `Iterator[T] => Iterator[Array[T]]`. Removed the extraneous `toSeq`,
though.
---
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]
