Revert "SQL style null-on-empty semantics for Sum and Mean" by joseph-isaacs · Pull Request #9324 · vortex-data/vortex

Performance Regression: -1.51%

⚡ 11 improved benchmarks
❌ 8 regressed benchmarks
✅ 1921 untouched benchmarks
⏩ 89 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

	Mode	Benchmark	`BASE`	`HEAD`	Efficiency
❌	Simulation	`list_sum_nullable_elements_medium`	4.7 ms	11.6 ms	-59.25%
❌	Simulation	`list_sum_nullable_elements_large`	464.5 ms	1,093.8 ms	-57.54%
❌	Simulation	`listview_sum_large`	150.5 ms	231.7 ms	-35.04%
❌	Simulation	`fsl_sum_medium`	1.2 ms	1.7 ms	-33.2%
❌	Simulation	`list_sum_large`	185.3 ms	275.4 ms	-32.7%
❌	Simulation	`fsl_sum_large`	96.5 ms	140 ms	-31.04%
❌	Simulation	`listview_sum_medium`	1.6 ms	2.1 ms	-23.35%
❌	Simulation	`list_sum_medium`	2 ms	2.5 ms	-21.37%
⚡	Simulation	`sum_f64_all_valid`	196.6 µs	88.1 µs	×2.2
⚡	Simulation	`sum_i32_nullable_all_valid`	188.5 µs	95.7 µs	+96.96%
⚡	Simulation	`fsl_sum_small`	213.4 µs	135.5 µs	+57.48%
⚡	Simulation	`sum_i32_clustered_nulls`	232.6 µs	151 µs	+54.05%
⚡	Simulation	`sum_f64_clustered_nulls`	241.3 µs	158.9 µs	+51.81%
⚡	Simulation	`listview_sum_small`	224.3 µs	176.2 µs	+27.28%
⚡	Simulation	`list_sum_small`	317.6 µs	279.4 µs	+13.7%
⚡	Simulation	`sparse_null_count`	86.3 µs	76.3 µs	+13.06%
⚡	Simulation	`search_index_in_range_chunked`	6.4 ms	5.8 ms	+10.83%
⚡	Simulation	`sparse_sum`	571.5 µs	517.5 µs	+10.43%
⚡	Simulation	`null_count_run_end[(10000, 1024, 0.5)]`	84.8 µs	77 µs	+10.03%

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

_{Comparing revert-9113-mk/unify-sums-v2 (00d28df) with develop (4b37237)}

89 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. ↩