zhengruifeng opened a new pull request, #58111: URL: https://github.com/apache/spark/pull/58111
### What changes were proposed in this pull request? This PR updates `MultivariateOnlineSummarizer` and `SummarizerBuffer` to propagate `NaN` values for `min` and `max` when a feature contains `NaN`. The add path now uses `math.min` and `math.max`, matching the merge path behavior and allowing `NaN` to be preserved in the running min/max arrays. ### Why are the changes needed? Previously, `min` and `max` were updated through direct `<` and `>` comparisons. Since comparisons with `NaN` are false, `NaN` values could be ignored and finite values could be returned even when a column contained `NaN`. ### Does this PR introduce _any_ user-facing change? Yes. `Statistics.colStats`, `MultivariateOnlineSummarizer`, and `Summarizer` now return `NaN` for `min` and `max` in a feature dimension if that dimension contains `NaN`. ### How was this patch tested? Added regression tests for both `MultivariateOnlineSummarizer` and `Summarizer`. Ran: ``` build/sbt "mllib/testOnly org.apache.spark.mllib.stat.MultivariateOnlineSummarizerSuite org.apache.spark.ml.stat.SummarizerSuite" ``` ### Was this patch authored or co-authored using generative AI tooling? Generated-by: Codex (GPT-5) -- 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]
