Github user kiszk commented on a diff in the pull request: https://github.com/apache/spark/pull/18747#discussion_r145148278 --- Diff: sql/core/src/main/scala/org/apache/spark/sql/execution/columnar/InMemoryRelation.scala --- @@ -63,6 +64,23 @@ case class InMemoryRelation( val batchStats: LongAccumulator = child.sqlContext.sparkContext.longAccumulator) extends logical.LeafNode with MultiInstanceRelation { + /** + * If true, get data from ColumnVector in ColumnarBatch, which are generally faster. + * If false, get data from UnsafeRow build from ColumnVector + */ + private[columnar] val useColumnarBatches: Boolean = { --- End diff -- You are right. Previous implementation required to change execution path in the write path, too. After review comments, this flag is not referenced in `InMemoryRelation`. I will move this flag to `InMemoryTableScanExec`.
--- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org