Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/engusmaze/frand

Blazingly Fast Pseudo Random Number Generator written in Rust
https://github.com/engusmaze/frand

no-std prng rand random random-generation rust rust-lang

Last synced: about 1 month ago
JSON representation

Blazingly Fast Pseudo Random Number Generator written in Rust

Awesome Lists containing this project

README

        

# FRand

[![Crate](https://img.shields.io/crates/v/frand.svg)](https://crates.io/crates/frand)

**FRand** is a blazingly fast, small, and simple pseudo-random number generator (PRNG) written in Rust. The advantage of using FRand is that it can produce more random numbers per second than other libraries. It also produces high-quality random numbers using a fast **non-cryptographic** hashing algorithm.

To find the best constants for the algorithm, I used an automated program that tried many random constants per second and measured the bias of the output. The bias was estimated using the avalanche effect, which is a property of good hash functions that ensures that a small change in the input produces a large change in the output. The program selected the constants that minimized the bias and maximized the randomness.

**FRand** is really simple to use. Here is a simple example of how to use FRand to generate a random float:

```rs
use frand::Rand;

let mut rng = Rand::new();
println!("{}", rng.gen::());
```

## Speeds

### u64

| rand::ThreadRng | rand::SmallRng | fastrand::Rng | frand::Rand |
| :-------------: | :------------: | :-----------: | :---------: |
| 1.00x | 3.74x | 5.80x | 6.09x |

### f64

| rand::ThreadRng | rand::SmallRng | fastrand::Rng | frand::Rand |
| :-------------: | :------------: | :-----------: | :---------: |
| 1.00x | 3.79x | 2.34x | 5.57x |

### u128

| rand::ThreadRng | rand::SmallRng | fastrand::Rng | frand::Rand |
| :-------------: | :------------: | :-----------: | :---------: |
| 1.00x | 2.15x | 6.79x | 7.24x |

If you have suggestions on how to improve this library, you can contribute to this project!