navaneethsujith09 commented on issue #4945:
URL: https://github.com/apache/gluten/issues/4945#issuecomment-4764025452

   @taiyang-li Thank you for reviewing my PR. I have removed the comment. 
However, I have a question regarding enabling the skewness UTs in Spark.
   
   From what I can tell, the only Spark SQL golden-file UT covering skewness is 
in group-by.sql, which is currently listed in IGNORE_SQL_QUERY_LIST in 
ClickHouseSQLQueryTestSettings.scala (Spark 3.5/4.0/4.1). I'm happy to re-open 
it similar to how #12033 re-opened other queries, but I wanted to confirm the 
expected approach first.
   
   The challenge is that the aggregate query in group-by.sql also uses 
KURTOSIS, which is still unsupported by the CH backend. As a result, that query 
would fall back to vanilla Spark even though skewness now executes natively. 
Since the golden-file tests compare query results rather than asserting 
no-fallback, I would expect the test to still pass, but enabling it would mean 
re-opening the entire group-by.sql file rather than only the skewness query.
   
   Would you prefer that I:
   (a) re-open the full group-by.sql file and rely on CI to validate the 
remaining queries, or
   (b) add a dedicated skewness SQL test so that only skewness is enabled?


-- 
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]

Reply via email to