Nexus Sort

A cognitive, multi-strategy sorting framework in Rust, designed for high performance by intelligently choosing the best sorting strategy for your data. This library was designed and built as a collaboration between a human user and a large language model (Google's Gemini).

Core Concept

Instead of a "one size fits all" approach, NexusSort analyzes the input data and selects the optimal algorithm from its arsenal:

For integers, it uses its Cognitive Engine to choose between a hyper-optimized stable sort (Timsort) or unstable sort (Introsort). For random data, it defaults to a Radix Sort for maximum performance.
For floats, it uses a specialized f32 sorter that can leverage a bit-level Radix Sort, yielding massive performance gains over traditional comparison sorts.

Performance: Proof of Victory

Benchmarks were conducted against the Rust 1.xx.x standard library's sorting functions. The results demonstrate significant performance gains.

Integer Sorting (Random `i32` Data)

NexusSort (using its Radix path) vs. std::sort_unstable.

Input Size	NexusSort (µs)	`std::sort_unstable` (µs)	Performance Gain
1,000	~12.5 µs	~24.1 µs	~93% Faster
10,000	~172.3 µs	~215.2 µs	~25% Faster

Float Sorting (Random `f32` Data)

NexusSort (using its Radix path) vs. std::sort (stable).

Input Size	NexusSort (µs)	`std::sort` (µs)	Performance Gain
1,000	~16.3 µs	~50.8 µs	~212% Faster
10,000	~174.4 µs	~1129.7 µs	~547% Faster

Usage

Add this to your Cargo.toml:

[dependencies]

nexus_sort_core = "1.0.1"

And use it in your code:

use nexus_sort_core::{nexus_sort, nexus_sort_f32};

// Sorting integers
let mut numbers = vec!;
nexus_sort(&mut numbers);

// Sorting floats
let mut floats = vec!;
nexus_sort_f32(&mut floats);

License

This project is licensed under the MIT License.

nexus_sort_core 1.0.1