Github user icexelloss commented on a diff in the pull request: https://github.com/apache/spark/pull/18732#discussion_r143741944 --- Diff: sql/core/src/main/scala/org/apache/spark/sql/execution/python/ArrowEvalPythonExec.scala --- @@ -44,14 +73,18 @@ 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 batchSize = conf.arrowMaxRecordsPerBatch + // DO NOT use iter.grouped(). See BatchIterator. + val batchIter = if (batchSize > 0) new BatchIterator(iter, batchSize) else Iterator(iter) + val columnarBatchIter = new ArrowPythonRunner( - funcs, conf.arrowMaxRecordsPerBatch, bufferSize, reuseWorker, + funcs, bufferSize, reuseWorker, PythonEvalType.SQL_PANDAS_UDF, argOffsets, schema) - .compute(iter, context.partitionId(), context) + .compute(batchIter, context.partitionId(), context) new Iterator[InternalRow] { - var currentIter = if (columnarBatchIter.hasNext) { + private var currentIter = if (columnarBatchIter.hasNext) { --- End diff -- I think so. The variable is reassigned for each columnar batch
--- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org