comphead commented on PR #2040:
URL:
https://github.com/apache/datafusion-comet/pull/2040#issuecomment-3111563959
Fixed windows and crypto tests, the `ignore nulls` correctness issue is
weird. Ive localized it with 3 rows test
```
test("first/last with ignore null") {
val data = Range(0, 3).flatMap(n => Seq((n, 1), (n, 2))).toDF("a", "b")
withTempDir { dir =>
val filename = s"${dir.getAbsolutePath}/first_last_ignore_null.parquet"
data.repartition(2).write.parquet(filename)
withSQLConf(CometConf.COMET_BATCH_SIZE.key -> "1") {
spark.read.parquet(filename).createOrReplaceTempView("t1")
for (expr <- Seq("first", "last")) {
// deterministic query that should return one non-null value per
group
val df = spark.sql(
s"SELECT a, $expr(IF(b==1,null,b)) IGNORE NULLS FROM t1 GROUP BY
a ORDER BY a")
checkSparkAnswerAndOperator(df)
}
}
}
}
```
It has something with Batch size and number of input files.
--
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]
