Github user setjet commented on a diff in the pull request:
https://github.com/apache/spark/pull/18113#discussion_r150392495
--- Diff:
sql/core/src/main/scala/org/apache/spark/sql/execution/aggregate/typedaggregators.scala
---
@@ -76,26 +76,126 @@ class TypedCount[IN](val f: IN => Any) extends
Aggregator[IN, Long, Long] {
// Java api support
def this(f: MapFunction[IN, Object]) = this((x: IN) =>
f.call(x).asInstanceOf[Any])
+
def toColumnJava: TypedColumn[IN, java.lang.Long] = {
toColumn.asInstanceOf[TypedColumn[IN, java.lang.Long]]
}
}
+class TypedAverage[IN](val f: IN => Double)
+ extends Aggregator[IN, (Double, Long), Double] {
-class TypedAverage[IN](val f: IN => Double) extends Aggregator[IN,
(Double, Long), Double] {
override def zero: (Double, Long) = (0.0, 0L)
override def reduce(b: (Double, Long), a: IN): (Double, Long) = (f(a) +
b._1, 1 + b._2)
- override def finish(reduction: (Double, Long)): Double = reduction._1 /
reduction._2
--- End diff --
Order of functions is consistent among all aggregation functions: zero,
reduce, merge finish. Hence the swap of location of the merge and finish
functions
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
