gene-bordegaray commented on code in PR #23239:
URL: https://github.com/apache/datafusion/pull/23239#discussion_r3505701743
##########
datafusion/core/tests/physical_optimizer/enforce_distribution.rs:
##########
@@ -700,6 +748,201 @@ impl TestConfig {
}
}
+#[test]
+fn range_aggregate_keeps_range_partitioning_at_subset_threshold() ->
Result<()> {
+ let input = parquet_exec_with_output_partitioning(range_partitioning(
+ "a",
+ [10, 20, 30],
+ SortOptions::default(),
+ )?);
+ let aggregate =
+ aggregate_exec_with_alias(input, vec![("a".to_string(),
"a".to_string())]);
+
+ let plan = TestConfig::default()
+ .with_query_execution_partitions(4)
+ .to_plan(aggregate, &DISTRIB_DISTRIB_SORT);
+
+ assert_plan!(
+ plan,
+ @r"
+ AggregateExec: mode=FinalPartitioned, gby=[a@0 as a], aggr=[]
+ AggregateExec: mode=Partial, gby=[a@0 as a], aggr=[]
+ DataSourceExec: file_groups={4 groups: [[p0], [p1], [p2], [p3]]},
projection=[a, b, c, d, e], output_partitioning=Range([a@0 ASC], [(10), (20),
(30)], 4), file_type=parquet
+ "
+ );
+
+ Ok(())
+}
+
+#[test]
+fn range_aggregate_keeps_range_subset_partitioning_at_subset_threshold() ->
Result<()> {
+ let input = parquet_exec_with_output_partitioning(range_partitioning(
+ "a",
+ [10, 20, 30],
+ SortOptions::default(),
+ )?);
+ let aggregate = aggregate_exec_with_alias(
+ input,
+ vec![
+ ("a".to_string(), "a".to_string()),
+ ("b".to_string(), "b".to_string()),
+ ],
+ );
+
+ let plan = TestConfig::default()
+ .with_query_execution_partitions(4)
+ .to_plan(aggregate, &DISTRIB_DISTRIB_SORT);
+
+ assert_plan!(
+ plan,
+ @r"
+ AggregateExec: mode=FinalPartitioned, gby=[a@0 as a, b@1 as b], aggr=[]
+ AggregateExec: mode=Partial, gby=[a@0 as a, b@1 as b], aggr=[]
+ DataSourceExec: file_groups={4 groups: [[p0], [p1], [p2], [p3]]},
projection=[a, b, c, d, e], output_partitioning=Range([a@0 ASC], [(10), (20),
(30)], 4), file_type=parquet
+ "
+ );
+
+ Ok(())
+}
+
+#[test]
+fn range_aggregate_rehashes_below_subset_threshold() -> Result<()> {
+ let input = parquet_exec_with_output_partitioning(range_partitioning(
+ "a",
+ [10, 20],
+ SortOptions::default(),
+ )?);
+ let aggregate =
+ aggregate_exec_with_alias(input, vec![("a".to_string(),
"a".to_string())]);
+
+ let plan = TestConfig::default()
+ .with_query_execution_partitions(4)
+ .to_plan(aggregate, &DISTRIB_DISTRIB_SORT);
+
+ assert_plan!(
+ plan,
+ @r"
+ AggregateExec: mode=FinalPartitioned, gby=[a@0 as a], aggr=[]
+ RepartitionExec: partitioning=Hash([a@0], 4), input_partitions=4
+ AggregateExec: mode=Partial, gby=[a@0 as a], aggr=[]
+ RepartitionExec: partitioning=RoundRobinBatch(4), input_partitions=3
+ DataSourceExec: file_groups={3 groups: [[p0], [p1], [p2]]},
projection=[a, b, c, d, e], output_partitioning=Range([a@0 ASC], [(10), (20)],
3), file_type=parquet
+ "
+ );
+
+ Ok(())
+}
+
+#[test]
+fn range_aggregate_rehashes_preserved_file_partitions_below_target() ->
Result<()> {
+ let input = parquet_exec_with_output_partitioning(range_partitioning(
+ "a",
+ [10, 20],
+ SortOptions::default(),
+ )?);
+ let aggregate =
+ aggregate_exec_with_alias(input, vec![("a".to_string(),
"a".to_string())]);
+
+ let plan = TestConfig::default()
+ .with_query_execution_partitions(4)
+ .with_preserve_file_partitions(1)
+ .to_plan(aggregate, &DISTRIB_DISTRIB_SORT);
+
+ assert_plan!(
+ plan,
+ @r"
+ AggregateExec: mode=FinalPartitioned, gby=[a@0 as a], aggr=[]
+ RepartitionExec: partitioning=Hash([a@0], 4), input_partitions=4
+ AggregateExec: mode=Partial, gby=[a@0 as a], aggr=[]
+ RepartitionExec: partitioning=RoundRobinBatch(4), input_partitions=3
+ DataSourceExec: file_groups={3 groups: [[p0], [p1], [p2]]},
projection=[a, b, c, d, e], output_partitioning=Range([a@0 ASC], [(10), (20)],
3), file_type=parquet
+ "
+ );
+
+ Ok(())
+}
+
+#[test]
+fn range_aggregate_rehashes_incompatible_partitioning() -> Result<()> {
+ let input = parquet_exec_with_output_partitioning(range_partitioning(
+ "a",
+ [10, 20],
+ SortOptions::default(),
+ )?);
+ let aggregate =
+ aggregate_exec_with_alias(input, vec![("b".to_string(),
"b".to_string())]);
+
+ let plan = TestConfig::default()
+ .with_query_execution_partitions(3)
+ .to_plan(aggregate, &DISTRIB_DISTRIB_SORT);
+
+ assert_plan!(
+ plan,
+ @r"
+ AggregateExec: mode=FinalPartitioned, gby=[b@0 as b], aggr=[]
+ RepartitionExec: partitioning=Hash([b@0], 3), input_partitions=3
+ AggregateExec: mode=Partial, gby=[b@1 as b], aggr=[]
+ DataSourceExec: file_groups={3 groups: [[p0], [p1], [p2]]},
projection=[a, b, c, d, e], output_partitioning=Range([a@0 ASC], [(10), (20)],
3), file_type=parquet
+ "
+ );
+
+ Ok(())
+}
+
+#[test]
+fn range_grouping_set_aggregate_rehashes_with_grouping_id() -> Result<()> {
Review Comment:
We cannot disable this, I just extracted is logic and thought it was
interesting there was no uinit tests for this since it is quite a weird edge
case. This was introduced in #19853 and without it the subset satisfaction can
create correctness errors 👍
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
