Expand description
prime-diffusion — Stochastic processes: Ornstein-Uhlenbeck and geometric Brownian motion.
All public functions are pure (LOAD + COMPUTE only). No &mut, no side effects,
no hidden state. Noise is either supplied externally (caller-provided standard normal w)
or generated deterministically from a threaded u32 seed via prime_random::prng_gaussian.
§Temporal Assembly Model
- LOAD — read parameters + state
- COMPUTE — stochastic update formula
- APPEND — return
(next_value, next_seed)as a tuple
STORE and JUMP do not exist here. Seeded variants thread the seed forward like
prime_random::prng_next — pure state machines.
§Included
ou_step— Ornstein-Uhlenbeck step (caller-supplied noise)ou_step_seeded— OU step with deterministic Gaussian noise from threaded seedgbm_step— Geometric Brownian motion step (caller-supplied noise)gbm_step_seeded— GBM step with deterministic noise from threaded seed
Functions§
- gbm_
step - Geometric Brownian motion step with caller-supplied noise.
- gbm_
step_ seeded - Geometric Brownian motion step with deterministic seeded noise.
- ou_step
- Ornstein-Uhlenbeck step with caller-supplied noise.
- ou_
step_ seeded - Ornstein-Uhlenbeck step with deterministic seeded noise.