Better fixed-width buffer filtering by robert3005 · Pull Request #8723 · vortex-data/vortex

Performance Gate Passed

⚡ 15 improved benchmarks
✅ 1655 untouched benchmarks
⏩ 44 skipped benchmarks¹

	Mode	Benchmark	`BASE`	`HEAD`	Efficiency
⚡	Simulation	`patterns_i128[Contiguous]`	234.6 µs	17.8 µs	×13
⚡	Simulation	`random_i64[0.8]`	277.6 µs	157.3 µs	+76.49%
⚡	Simulation	`random_i64[0.5]`	204.5 µs	138.3 µs	+47.93%
⚡	Simulation	`random_i32[0.8]`	128.4 µs	87 µs	+47.56%
⚡	Simulation	`random_i32[0.5]`	114.2 µs	78.7 µs	+45.04%
⚡	Simulation	`random_i128[0.8]`	389 µs	276.2 µs	+40.86%
⚡	Simulation	`patterns_i128[Runs]`	244.8 µs	173.8 µs	+40.8%
⚡	Simulation	`random_i128[0.5]`	290.1 µs	218.7 µs	+32.67%
⚡	Simulation	`patterns_i128[Random]`	290.1 µs	219.2 µs	+32.34%
⚡	Simulation	`random_i256[0.8]`	639.3 µs	525.7 µs	+21.61%
⚡	Simulation	`random_i32[0.95]`	108.6 µs	89.4 µs	+21.37%
⚡	Simulation	`random_i256[0.5]`	445.4 µs	373.9 µs	+19.14%
⚡	Simulation	`take_filter_primitive_large_random_mask_random_indices[(12500, 25000)]`	366.9 µs	312.4 µs	+17.46%
⚡	Simulation	`random_i256[0.01]`	31.8 µs	28.6 µs	+11.5%
⚡	Simulation	`random_i64[0.95]`	182.6 µs	165.5 µs	+10.33%

Tip

Curious why this is faster? Comment @codspeedbot explain why this is faster on this PR, or directly use the CodSpeed MCP with your agent.

_{Comparing rk/filterbuffer (be5e458) with develop (05c7c2f)}

44 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. ↩