Struct ContinuousTimeMarkov 

pub struct ContinuousTimeMarkov {
    pub q: Vec<f64>,
    pub n_states: usize,
}

Continuous-time Markov chain with infinitesimal generator matrix Q.

Fields

q: Vec<f64>

Generator matrix Q (n x n). Rows sum to zero; off-diagonals >= 0.

n_states: usize

Number of states.

Implementations

impl ContinuousTimeMarkov

pub fn new(q: Vec<f64>, n_states: usize) -> Self

Create a CTMC from generator matrix Q.

pub fn simulate( &self, init: usize, t_max: f64, seed: u64, ) -> (Vec<f64>, Vec<usize>)

Simulate a CTMC trajectory up to time T using Gillespie’s algorithm.

Returns (times, states) pairs representing the jump times and new states.

pub fn kolmogorov_forward(&self, p0: &[f64], t: f64) -> Vec<f64>

Solve the Kolmogorov forward equations dp/dt = p * Q at time T.

Uses matrix exponentiation: p(T) = p(0) * exp(Q * T).

pub fn stationary_distribution(&self, tol: f64, max_iter: usize) -> Vec<f64>

Compute the stationary distribution pi satisfying pi Q = 0, sum(pi) = 1.

pub fn mean_first_passage_time(&self, from: usize, to: usize) -> f64

Compute the mean first-passage time from state i to state j.

Uses the fundamental matrix method for ergodic CTMCs.

pub fn mean_sojourn_time(&self, state: usize) -> f64

Compute the expected holding time in state i (mean sojourn time).