Github user kiszk commented on a diff in the pull request: https://github.com/apache/spark/pull/21102#discussion_r203319710 --- Diff: core/src/main/scala/org/apache/spark/util/collection/OpenHashSet.scala --- @@ -163,7 +187,7 @@ class OpenHashSet[@specialized(Long, Int) T: ClassTag]( * to a new position (in the new data array). */ def rehashIfNeeded(k: T, allocateFunc: (Int) => Unit, moveFunc: (Int, Int) => Unit) { - if (_size > _growThreshold) { + if (_occupied > _growThreshold) { --- End diff -- When 'remove' is called, '_size' is decremented. But, an entry is not released. This is a motivation to introduce 'occupied'. I will try to use another implementation without 'remove' while it may introduce some overhead.
--- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org