Github user srowen commented on a diff in the pull request:
https://github.com/apache/spark/pull/5075#discussion_r26662472
--- Diff: core/src/main/scala/org/apache/spark/rdd/PairRDDFunctions.scala
---
@@ -163,6 +163,28 @@ class PairRDDFunctions[K, V](self: RDD[(K, V)])
}
/**
+ * Returns the top k (largest) elements for each key from this RDD as
defined by the specified
+ * implicit Ordering[T].
+ * If the number of elements for a certain key is less than k, all of
them will be returned.
+ *
+ * @param num k, the number of top elements to return
+ * @param ord the implicit ordering for T
+ * @return an RDD that contains the top k values for each key
+ */
+ def topByKey(num: Int)(implicit ord: Ordering[V]): RDD[(K, Array[V])] = {
+ aggregateByKey(new BoundedPriorityQueue[V](num)(ord))(
+ seqOp = (queue, item) => {
+ queue += item
+ queue
+ },
+ combOp = (queue1, queue2) => {
+ queue1 ++= queue2
+ queue1
+ }
+ ).mapValues(_.toArray.sorted(ord.reverse))
--- End diff --
Does the `toArray` not already give you the top k in order? this seems to
be the behavior already as it returns the array formed from the iterator in the
underlying `PriorityQueue`. Worth testing I think. (That said, I suppose
sorting an already-sorted array is pretty fast.)
Nit: a few lines above, an extra space in front of `=>`
---
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]
