haohuaijin opened a new issue, #19638:
URL: https://github.com/apache/datafusion/issues/19638

   ### Is your feature request related to a problem or challenge?
   
   current the `GroupedTopKAggregateStream` support two type of query
   ```sql
   select id, max(time) from t group by id order by max(time) desc limit 10
   select id, min(time) from t group by id order by min(time) asc limit 10
   ```
   we have another use case that i find it also can use 
GroupedTopKAggregateStream to spped up, like below query
   ```sql
   select distinct id from t order by id desc/asc limit 10
   select id from t group by id order by id desc/asc limit 10
   ```
   because If a certain `id = x` is in the global top 10(the second phase of 
AggregateExec), then `x` must appear in the local top 10(the first phase of 
AggregateExec) of at least one partition.
   
   ### Describe the solution you'd like
   
   1. modify the `TopKAggregation` optimizer rule to pass the information to 
AggregateExec
   2. modify `GroupedTopKAggregateStream` to support this case
   
   ### Describe alternatives you've considered
   
   _No response_
   
   ### Additional context
   
   _No response_


-- 
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