Github user gatorsmile commented on the issue:
https://github.com/apache/spark/pull/16404
```Scala
sql("select a + rand() from testData2 group by a, a + rand()").explain(true)
```
After we merging this PR, I am afraid we might hitting a common
misunderstanding. Users might assume `a + rand()` in the aggregate and group by
are referring to the same expression. However, the non-deterministic
expressions `rand()` in aggregate and group by actually return different
results. See the plan
```
(CAST(a AS DOUBLE) + rand(-8319334304660323190)): double
Aggregate [a#23, (cast(a#23 as double) + _nondeterministic#249)],
[(cast(a#23 as double) + rand(-8319334304660323190)) AS (CAST(a AS DOUBLE) +
rand(-8319334304660323190))#248]
+- Project [a#23, b#24, rand(-993354445377164024) AS _nondeterministic#249]
+- SubqueryAlias testdata2, `testData2`
+- SerializeFromObject [assertnotnull(assertnotnull(input[0,
org.apache.spark.sql.test.SQLTestData$TestData2, true], top level Product input
object), - root class: "org.apache.spark.sql.test.SQLTestData.TestData2").a AS
a#23, assertnotnull(assertnotnull(input[0,
org.apache.spark.sql.test.SQLTestData$TestData2, true], top level Product input
object), - root class: "org.apache.spark.sql.test.SQLTestData.TestData2").b AS
b#24]
+- ExternalRDD [obj#22]
```
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
