Repository navigation
[EPIC] Improvements to GroupColumn multi-column aggregation performance #12680
Description
Is your feature request related to a problem or challenge?
In #12269 @jayzhan211 made significant improvements to how group values are stored in multi-column aggregations. There are a few follow ups to this work that I wanted to track here
Describe the solution you'd like
- Minor: Encapsulate type check in GroupValuesColumn, avoid panic #12620
- Rename aggregation modules, GroupColumn #12619
- docs: improve the documentation for Aggregate code #12617
- Refactor PrimitiveGroupValueBuilder to use
MaybeNullBufferBuilder#12623 - Reduce code duplication in
PrimitiveGroupValueBuilderwith const generics #12703 - Refactor
ByteGroupValueBuilderto useMaybeNullBufferBuilder#12681 - fix
equal_toinPrimitiveGroupValueBuilder#12758 - Fix
equal_toinByteGroupValueBuilder#12770 - potentially and optimize the take_n for the boolean builder using arrow-rs (would help streaming aggregates)
- Implement special Groups for StringViews #12771
- Implement GroupColumn Decimal128Array #13505
- Support vectorized append and compare for multi group by #12996
- Reorganize the GroupColumn implementations into more coherent modules #13262
- Implement Specialized
GroupColumnforDate/Time/Timestamptypes for multi-columnGROUP BY#13263
Describe alternatives you've considered
No response
Additional context
No response
Activity
[-]Improvements to GroupColumn multi-column aggregation performance[/-][+][EPIC] Improvements to GroupColumn multi-column aggregation performance[/+]alamb commentedon Nov 4, 2024
@Rachelint has a special vectorized comparison PR here as well: #12996
alamb commentedon Dec 30, 2024
I believe all the currently known items for this epic are completed so closing this issue for now. Let's open new issues if/when needed