This is an automated email from the ASF dual-hosted git repository.
wenchen pushed a commit to branch branch-3.4
in repository https://gitbox.apache.org/repos/asf/spark.git
The following commit(s) were added to refs/heads/branch-3.4 by this push:
new 5e9ad2e71f9 [SPARK-42259][SQL] ResolveGroupingAnalytics should take
care of Python UDAF
5e9ad2e71f9 is described below
commit 5e9ad2e71f9a22d2f1b2d47b2e6bee323b354eb7
Author: Wenchen Fan <wenc...@databricks.com>
AuthorDate: Wed Feb 1 17:36:14 2023 +0800
[SPARK-42259][SQL] ResolveGroupingAnalytics should take care of Python UDAF
### What changes were proposed in this pull request?
This is a long-standing correctness issue with Python UDAF and grouping
analytics. The rule `ResolveGroupingAnalytics` should take care of Python UDAF
when matching aggregate expressions.
### Why are the changes needed?
bug fix
### Does this PR introduce _any_ user-facing change?
Yes, the query result was wrong before
### How was this patch tested?
existing tests
Closes #39824 from cloud-fan/python.
Authored-by: Wenchen Fan <wenc...@databricks.com>
Signed-off-by: Wenchen Fan <wenc...@databricks.com>
(cherry picked from commit 1219c8492376e038894111cd5d922229260482e7)
Signed-off-by: Wenchen Fan <wenc...@databricks.com>
---
.../spark/sql/catalyst/analysis/Analyzer.scala | 2 +-
.../results/udaf/udaf-group-analytics.sql.out | 58 +++++++++++-----------
2 files changed, 30 insertions(+), 30 deletions(-)
diff --git
a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala
b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala
index 48ea0460725..7a92c46577d 100644
---
a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala
+++
b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala
@@ -622,7 +622,7 @@ class Analyzer(override val catalogManager: CatalogManager)
// AggregateExpression should be computed on the unmodified value of
its argument
// expressions, so we should not replace any references to grouping
expression
// inside it.
- case e: AggregateExpression =>
+ case e if AggregateExpression.isAggregate(e) =>
aggsBuffer += e
e
case e if isPartOfAggregation(e) => e
diff --git
a/sql/core/src/test/resources/sql-tests/results/udaf/udaf-group-analytics.sql.out
b/sql/core/src/test/resources/sql-tests/results/udaf/udaf-group-analytics.sql.out
index b8c94b19d81..f0be6f43642 100644
---
a/sql/core/src/test/resources/sql-tests/results/udaf/udaf-group-analytics.sql.out
+++
b/sql/core/src/test/resources/sql-tests/results/udaf/udaf-group-analytics.sql.out
@@ -15,18 +15,18 @@ SELECT a + b, b, udaf(a - b) FROM testData GROUP BY a + b,
b WITH CUBE
struct<(a + b):int,b:int,udaf((a - b)):int>
-- !query output
2 1 1
-2 NULL 0
+2 NULL 1
3 1 1
3 2 1
-3 NULL 0
+3 NULL 2
4 1 1
4 2 1
-4 NULL 0
+4 NULL 2
5 2 1
-5 NULL 0
+5 NULL 1
NULL 1 3
NULL 2 3
-NULL NULL 0
+NULL NULL 6
-- !query
@@ -36,16 +36,16 @@ struct<a:int,b:int,udaf(b):int>
-- !query output
1 1 1
1 2 1
-1 NULL 0
+1 NULL 2
2 1 1
2 2 1
-2 NULL 0
+2 NULL 2
3 1 1
3 2 1
-3 NULL 0
+3 NULL 2
NULL 1 3
NULL 2 3
-NULL NULL 0
+NULL NULL 6
-- !query
@@ -54,16 +54,16 @@ SELECT a + b, b, udaf(a - b) FROM testData GROUP BY a + b,
b WITH ROLLUP
struct<(a + b):int,b:int,udaf((a - b)):int>
-- !query output
2 1 1
-2 NULL 0
+2 NULL 1
3 1 1
3 2 1
-3 NULL 0
+3 NULL 2
4 1 1
4 2 1
-4 NULL 0
+4 NULL 2
5 2 1
-5 NULL 0
-NULL NULL 0
+5 NULL 1
+NULL NULL 6
-- !query
@@ -73,14 +73,14 @@ struct<a:int,b:int,udaf(b):int>
-- !query output
1 1 1
1 2 1
-1 NULL 0
+1 NULL 2
2 1 1
2 2 1
-2 NULL 0
+2 NULL 2
3 1 1
3 2 1
-3 NULL 0
-NULL NULL 0
+3 NULL 2
+NULL NULL 6
-- !query
@@ -416,14 +416,14 @@ GROUP BY course, earnings GROUPING SETS((), (course),
(course, earnings)) ORDER
-- !query schema
struct<course:string,sum:int>
-- !query output
-NULL 0
-Java 0
+NULL 5
Java 1
Java 1
-dotNET 0
+Java 2
dotNET 1
dotNET 1
dotNET 1
+dotNET 3
-- !query
@@ -432,14 +432,14 @@ GROUP BY course, earnings GROUPING SETS((), (course),
(course, earnings)) ORDER
-- !query schema
struct<course:string,sum:int,grouping_id(course, earnings):bigint>
-- !query output
-NULL 0 3
-Java 0 1
+NULL 5 3
Java 1 0
Java 1 0
-dotNET 0 1
+Java 2 1
dotNET 1 0
dotNET 1 0
dotNET 1 0
+dotNET 3 1
-- !query
@@ -468,16 +468,16 @@ SELECT a + b AS k, b, udaf(a - b) FROM testData GROUP BY
ROLLUP(k, b)
struct<k:int,b:int,udaf((a - b)):int>
-- !query output
2 1 1
-2 NULL 0
+2 NULL 1
3 1 1
3 2 1
-3 NULL 0
+3 NULL 2
4 1 1
4 2 1
-4 NULL 0
+4 NULL 2
5 2 1
-5 NULL 0
-NULL NULL 0
+5 NULL 1
+NULL NULL 6
-- !query
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@spark.apache.org
For additional commands, e-mail: commits-h...@spark.apache.org
