Aggregate Fns: Mean

[gatesn]

Summary

This PR implements a new Mean aggregate function for computing the arithmetic mean of numeric values in Vortex arrays.

The implementation includes:

• A new Mean struct that implements the AggregateFnVTable trait
• A MeanAccumulator that maintains running sum and count statistics
• Support for both integer and floating-point input types, with f64 output
• Proper handling of null values through validity masks
• State management for distributed aggregation scenarios through merge operations
• Automatic registration of the Mean function in the default AggregateFnSession

The accumulator efficiently processes batches by converting all numeric types to f64 and tracking valid values based on the array's validity mask. It supports flushing intermediate results for grouped operations and merging state from other accumulators for distributed computation.

Testing

The implementation includes comprehensive unit tests covering:

• Basic mean calculation for i32 and f64 arrays
• Null value handling (partial nulls and all-null arrays)
• Empty flush scenarios
• Multi-group aggregation with flush operations
• State merging functionality for distributed aggregation

All tests verify correct f64 output values and proper null handling in edge cases.

[gatesn]

Warning

This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more

• Aggregate Fns: Sum #6723
• Aggregate Fns: Mean #6722 👈 (View in Graphite)
• Aggregate Fns #6721
• develop

This stack of pull requests is managed by Graphite. Learn more about stacking.

[codspeed-hq]

Merging this PR will degrade performance by 18.01%

❌ 3 regressed benchmarks
✅ 951 untouched benchmarks
⏩ 1466 skipped benchmarks¹

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	bench_many_nulls[0.9]	481.8 µs	547.1 µs	-11.94%
❌	Simulation	take_map[(0.1, 0.1)]	916.5 µs	1,032.1 µs	-11.21%
❌	Simulation	take_map[(0.1, 1.0)]	3.6 ms	4.4 ms	-18.01%

---

_{Comparing ngates/aggregate-fns-mean (1f43f0b) with ngates/aggregate-fns (6d9bcac)}

Footnotes

1. 1466 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩