abhishekrb19 opened a new pull request #10211: URL: https://github.com/apache/druid/pull/10211
This patch adds vectorization support for the `longMin` aggregator to speed up queries that use the `MIN` aggregator. Summary of changes: - Added a new class `LongMinVectorAggregator` that implements the `VectorAggregator` interface. - Modified some existing test cases to include this vectorized in `SchemaEvolutionTest`. - Added a new simple test case to `CalciteQueryTest` and modified an existing one to demonstrate that vectorization _does_ work for the `longMin` aggregator. - Enabled null handling in the `LongMinAggregatorTest` to test both null-compatibility and non null-compatibility code paths. - Query context documentation update. - Miscellaneous minor tweaks to java doc comments. Once this is reviewed, I am hoping to also do the other min and max aggregators for float/double/long types, so we can trim down the list of aggregators that do not support vectorization :). Also, please suggest if I could add more tests or modify existing ones to ensure more code coverage, so the new code paths are exercised. Thanks! /cc: @gianm @jihoonson ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
