zhengruifeng opened a new pull request, #57483: URL: https://github.com/apache/spark/pull/57483
### What changes were proposed in this pull request? Makes `VectorIndexerModel.partialFeatureAttributes` lazy and adds a regression test showing that a high-dimensional model does not eagerly retain the derived attribute array. ### Why are the changes needed? The attribute array is derived state used only while preparing output metadata. Constructing it eagerly allocates one attribute per feature when a model is fitted or loaded, even when it is never transformed. Deferring the allocation avoids that unnecessary retained state. ### Does this PR introduce _any_ user-facing change? No. ### How was this patch tested? Added a `VectorIndexerSuite` regression test using a high-dimensional model whose eager derived attributes would exceed the expected size bound. The test suite was not run locally. ### Was this patch authored or co-authored using generative AI tooling? Generated-by: Codex -- 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]
