Github user ueshin commented on a diff in the pull request:
https://github.com/apache/spark/pull/18732#discussion_r141829344
--- Diff:
sql/core/src/main/scala/org/apache/spark/sql/execution/python/ArrowEvalPythonExec.scala
---
@@ -44,14 +44,17 @@ case class ArrowEvalPythonExec(udfs: Seq[PythonUDF],
output: Seq[Attribute], chi
val schemaOut =
StructType.fromAttributes(output.drop(child.output.length).zipWithIndex
.map { case (attr, i) => attr.withName(s"_$i") })
+ val batchedIter: Iterator[Iterator[InternalRow]] =
+ iter.grouped(conf.arrowMaxRecordsPerBatch).map(_.iterator)
+
--- End diff --
I guess this is for making `ArrowPythonRunner` reusable between current
pandas udf and `apply()` by taking `Iterator[Iterator[InternalRow]]` instead of
`Iterator[InternalRow]` as its input. The rows in grouped iterator will be one
`RecordBatch`.
I'm not sure whether it's good or not, though.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
