pepijnve commented on PR #16319:
URL: https://github.com/apache/datafusion/pull/16319#issuecomment-2956628323

   Looking at the `clickbench_partitioned` outliers. Wrt the code changes in 
this PR they seem pretty similar yet one has basically the opposite result of 
the other. What's interesting is that total time is actually lower for the run.
   
   Query 13 1.13x slower
   ```
   SELECT "SearchPhrase", COUNT(DISTINCT "UserID") AS u FROM 'hits.parquet' 
WHERE "SearchPhrase" <> '' GROUP BY "SearchPhrase" ORDER BY u DESC LIMIT 10;
   SortExec: TopK(fetch=10), expr=[u@1 DESC], preserve_partitioning=[false]
     ProjectionExec: expr=[SearchPhrase@0 as SearchPhrase, count(alias1)@1 as u]
       AggregateExec: mode=Single, gby=[SearchPhrase@0 as SearchPhrase], 
aggr=[count(alias1)]
         AggregateExec: mode=Single, gby=[SearchPhrase@1 as SearchPhrase, 
UserID@0 as alias1], aggr=[]
           CoalesceBatchesExec: target_batch_size=8192
             FilterExec: SearchPhrase@1 !=
               DataSourceExec: file_groups={1 group: 
[[Users/pepijn/RustroverProjects/datafusion/benchmarks/data/hits.parquet:0..14779976446]]},
 projection=[UserID, SearchPhrase], file_type=parquet, 
predicate=SearchPhrase@39 != , pruning_predicate=SearchPhrase_null_count@2 != 
row_count@3 AND (SearchPhrase_min@0 !=  OR  != SearchPhrase_max@1), 
required_guarantees=[SearchPhrase not in ()]
   ```
   
   Query 28: 1.10x faster
   ```
   SELECT REGEXP_REPLACE("Referer", '^https?://(?:www\.)?([^/]+)/.*$', '\1') AS 
k, AVG(length("Referer")) AS l, COUNT(*) AS c, MIN("Referer") FROM 
'hits.parquet' WHERE "Referer" <> '' GROUP BY k HAVING COUNT(*) > 100000 ORDER 
BY l DESC LIMIT 25;
   SortExec: TopK(fetch=25), expr=[l@1 DESC], preserve_partitioning=[false]
     ProjectionExec: 
expr=[regexp_replace(hits.parquet.Referer,Utf8("^https?://(?:www\.)?([^/]+)/.*$"),Utf8("\1"))@0
 as k, avg(character_length(hits.parquet.Referer))@1 as l, count(Int64(1))@2 as 
c, min(hits.parquet.Referer)@3 as min(hits.parquet.Referer)]
       CoalesceBatchesExec: target_batch_size=8192
         FilterExec: count(Int64(1))@2 > 100000
           AggregateExec: mode=Single, gby=[regexp_replace(Referer@0, 
^https?://(?:www\.)?([^/]+)/.*$, \1) as 
regexp_replace(hits.parquet.Referer,Utf8("^https?://(?:www\.)?([^/]+)/.*$"),Utf8("\1"))],
 aggr=[avg(character_length(hits.parquet.Referer)), count(Int64(1)), 
min(hits.parquet.Referer)]
             CoalesceBatchesExec: target_batch_size=8192
               FilterExec: Referer@0 !=
                 DataSourceExec: file_groups={1 group: 
[[Users/pepijn/RustroverProjects/datafusion/benchmarks/data/hits.parquet:0..14779976446]]},
 projection=[Referer], file_type=parquet, predicate=Referer@14 != , 
pruning_predicate=Referer_null_count@2 != row_count@3 AND (Referer_min@0 !=  OR 
 != Referer_max@1), required_guarantees=[Referer not in ()]
   ```


-- 
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: github-unsubscr...@datafusion.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscr...@datafusion.apache.org
For additional commands, e-mail: github-h...@datafusion.apache.org

Reply via email to