# partial_sort [![](https://img.shields.io/crates/v/partial_sort.svg)](https://crates.io/crates/partial_sort) [![](https://img.shields.io/crates/d/partial_sort.svg)](https://crates.io/crates/partial_sort) [![](https://docs.rs/partial_sort/badge.svg)](https://docs.rs/partial_sort/) [![](https://github.com/sundy-li/partial_sort/actions/workflows/Build.yml/badge.svg)](https://github.com/sundy-li/partial_sort/actions/workflows/Build.yml) partial_sort is Rust version of [std::partial_sort](https://en.cppreference.com/w/cpp/algorithm/partial_sort) ## Usage ```rust use partial_sort::PartialSort; fn main() { let mut vec = vec![4, 4, 3, 3, 1, 1, 2, 2]; vec.partial_sort(4, |a, b| a.cmp(b)); assert_eq!(&vec[0..4], &[1, 1, 2, 2]); } ``` ## Benches First we compare what happens when sorting the entire vector. Bench env: ``` $ uname -a Linux arch 6.8.7-arch1-1 #1 SMP PREEMPT_DYNAMIC Wed, 17 Apr 2024 15:20:28 +0000 x86_64 GNU/Linux $ cat /proc/cpuinfo | grep '\-Core' | head -1 model name : AMD Ryzen 9 5950X 16-Core Processor ``` Bench results: ``` nth_select sort 10000 limit 20 time: [8.6016 µs 8.6123 µs 8.6236 µs] partial sort 10000 limit 20 time: [5.3522 µs 5.3565 µs 5.3610 µs] // 1.6x faster partial sort 10000 limit 200 time: [15.736 µs 15.749 µs 15.763 µs] partial sort 10000 limit 2000 time: [111.97 µs 112.07 µs 112.18 µs] partial sort 10000 limit 10000 time: [219.21 µs 222.58 µs 225.85 µs] stdsort 10000 time: [81.795 µs 82.108 µs 82.383 µs] unstable stdsort 10000 time: [65.339 µs 65.355 µs 65.373 µs] heapsort 10000 time: [241.86 µs 242.09 µs 242.29 µs] partial reverse sort 10000 limit 20 time: [5.4574 µs 5.4696 µs 5.4843 µs] stdsort reverse 10000 time: [82.680 µs 82.751 µs 82.822 µs] ``` ## License Licensed under either of * Apache License, Version 2.0 ([LICENSE-APACHE](LICENSE-APACHE) or http://www.apache.org/licenses/LICENSE-2.0) * MIT license ([LICENSE-MIT](LICENSE-MIT) or http://opensource.org/licenses/MIT)