suibianwanwank opened a new pull request, #16234: URL: https://github.com/apache/datafusion/pull/16234
## Which issue does this PR close? - Closes #. ## Rationale for this change For unbounded aggregate window functions, the result is the same for all rows within a partition. Therefore, we can avoid repeatedly computing the window bounds and evaluating the function for each row, which can improve performance. Related Work in DuckDB：https://github.com/duckdb/duckdb/blob/main/src/function/window/window_constant_aggregator.cpp Benchmark results on h2o_small_window (run locally) show performance improvements for Q1 and Q3, with no noticeable changes in other queries. ### Before Optimization ``` Q1: -- Basic Window SELECT id1, id2, id3, v2, sum(v2) OVER () AS window_basic FROM large; Query 1 iteration 1 took 1297.6 ms and returned 10000000 rows Query 1 iteration 2 took 804.3 ms and returned 10000000 rows Query 1 iteration 3 took 582.2 ms and returned 10000000 rows Query 1 avg time: 894.72 ms Q3: -- PARTITION BY SELECT id1, id2, id3, v2, sum(v2) OVER (PARTITION BY id1) AS sum_by_id1, sum(v2) OVER (PARTITION BY id2) AS sum_by_id2, sum(v2) OVER (PARTITION BY id3) AS sum_by_id3 FROM large; Query 3 iteration 1 took 1532.6 ms and returned 10000000 rows Query 3 iteration 2 took 1755.2 ms and returned 10000000 rows Query 3 iteration 3 took 1821.4 ms and returned 10000000 rows Query 3 avg time: 1703.09 ms ``` ### After Optimization ``` Q1: -- Basic Window SELECT id1, id2, id3, v2, sum(v2) OVER () AS window_basic FROM large; Query 1 iteration 1 took 253.6 ms and returned 10000000 rows Query 1 iteration 2 took 190.3 ms and returned 10000000 rows Query 1 iteration 3 took 88.7 ms and returned 10000000 rows Query 1 avg time: 177.54 ms Q3: -- PARTITION BY SELECT id1, id2, id3, v2, sum(v2) OVER (PARTITION BY id1) AS sum_by_id1, sum(v2) OVER (PARTITION BY id2) AS sum_by_id2, sum(v2) OVER (PARTITION BY id3) AS sum_by_id3 FROM large; Query 3 iteration 1 took 1534.7 ms and returned 10000000 rows Query 3 iteration 2 took 1541.5 ms and returned 10000000 rows Query 3 iteration 3 took 1320.6 ms and returned 10000000 rows Query 3 avg time: 1465.59 ms ``` ## What changes are included in this PR? <!-- There is no need to duplicate the description in the issue here but it is sometimes worth providing a summary of the individual changes in this PR. --> ## Are these changes tested? <!-- We typically require tests for all PRs in order to: 1. Prevent the code from being accidentally broken by subsequent changes 2. Serve as another way to document the expected behavior of the code If tests are not included in your PR, please explain why (for example, are they covered by existing tests)? --> ## Are there any user-facing changes? <!-- If there are user-facing changes then we may require documentation to be updated before approving the PR. --> <!-- If there are any breaking changes to public APIs, please add the `api change` label. --> -- 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
