prime-diffusion

Stochastic diffusion processes — Ornstein-Uhlenbeck and Geometric Brownian Motion, with seeded and unseeded variants.

Part of the prime math ecosystem.

What's inside

• ou_step / ou_step_seeded — Ornstein-Uhlenbeck mean-reverting process
• gbm_step / gbm_step_seeded — Geometric Brownian Motion

Usage

use prime_diffusion::{ou_step_seeded, gbm_step_seeded};

// Ornstein-Uhlenbeck — mean-reverting random walk (useful for AI behavior, price simulation)
// Returns (next_value, next_seed)
let (x_next, seed_next) = ou_step_seeded(x, mu, theta, sigma, dt, seed);

// Geometric Brownian Motion — multiplicative random walk (asset prices, population growth)
let (x_next, seed_next) = gbm_step_seeded(x, mu, sigma, dt, seed);

Design

Seeded variants use prime-random to thread noise deterministically. Unseeded variants accept a pre-sampled Wiener increment w for integration with external noise sources.

License

MIT