helloitsheqing opened a new pull request, #44969: URL: https://github.com/apache/arrow/pull/44969
[Original Issue: #32381](https://github.com/apache/arrow/issues/32381) ### Description This pull request introduces a comprehensive test suite for the hash aggregation functionality in Apache Arrow’s compute kernels. The new test suite covers the following key components: **GroupedCountImpl:** - Validates merging of grouped counts with valid inputs. - Tests behavior with invalid group ID mappings (e.g., mismatched sizes, out-of-bounds indices). **GroupedMinMaxImpl:** - Verifies correctness of merging minimum and maximum values across groups. - Includes edge cases for invalid and out-of-bounds group ID mappings. **GroupedCountAllImpl:** - Tests merging of aggregated counts for "count all" scenarios. - Handles cases with valid and invalid inputs, ensuring robust error handling. The test suite is structured following the conventions of aggregate_test.cc, providing comprehensive coverage of valid and edge-case scenarios. ### Key Changes **New Test File:** Added hash_aggregate_test.cc in cpp/src/arrow/compute/kernels/. **CMake Configuration:** Updated cpp/src/arrow/compute/kernels/CMakeLists.txt to include the new test file. **Test Coverage:** Each test focuses on verifying correctness, error handling, and robustness for the Merge methods of: - GroupedCountImpl - GroupedMinMaxImpl - GroupedCountAllImpl ### Future Work If this test suite is approved, I plan to extend similar coverage to the remaining hash aggregation classes, including: - GroupedTDigestImpl - GroupedFirstLastImpl - GroupedBooleanAggregator - Other GroupedAggregator subclasses. This phased approach allows focused and iterative improvements to the test coverage for Arrow compute kernels. -- 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]
