manuzhang commented on PR #4792: URL: https://github.com/apache/datafusion-comet/pull/4792#issuecomment-4910085559
I ran `CometAggregateExpressionBenchmark` before/after on the same machine. Command: ```sh SPARK_GENERATE_BENCHMARK_FILES=1 make benchmark-org.apache.spark.sql.benchmark.CometAggregateExpressionBenchmark PROFILES="-Pjdk17 -Drat.skip=true" ``` Comet best-time comparison for the percentile cases: case | before | after | change -- | -- | -- | -- percentile_int_median | 255 ms | 259 ms | +1.6% percentile_long_median | 951 ms | 897 ms | -5.7% percentile_double_median | 259 ms | 266 ms | +2.7% percentile_double_p90 | 259 ms | 271 ms | +4.6% percentile_double_global | 98 ms | 103 ms | +5.1% percentile_double_high_card | 586 ms | 569 ms | -2.9% So I do not see a meaningful regression from the Spark-compatible percentile path. -- 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]
