Enum DurationDistrib 

pub enum DurationDistrib {
    Poisson {
        lambda: f64,
    },
    Geometric {
        p: f64,
    },
    Histogram {
        probs: Vec<f64>,
    },
}

Duration distribution for a single state.

All distributions are truncated to max_dur in the forward-backward algorithm; max_dur is passed as an argument to prob / log_prob to facilitate normalisation where needed.

Variants

Poisson

Poisson duration: P(dur = τ) ∝ λ^τ exp(−λ) / τ!, τ ≥ 1. Truncated and renormalised over τ = 1..max_dur.

Fields

lambda: f64

Geometric

Geometric: P(dur = τ) = p · (1−p)^{τ−1}, τ ≥ 1.

Fields

p: f64

Histogram

Histogram: probs[τ−1] = P(dur = τ), τ = 1..len.

Fields

probs: Vec<f64>

Implementations

impl DurationDistrib

pub fn prob(&self, tau: usize, max_dur: usize) -> f64

Probability of duration tau (≥ 1).

For Poisson, the distribution is truncated and renormalised to the range τ = 1..max_dur so that probabilities sum to 1 within the algorithm’s horizon.

pub fn log_prob(&self, tau: usize, max_dur: usize) -> f64

Log probability of duration tau; returns −∞ if prob = 0.

Trait Implementations

impl Clone for DurationDistrib

fn clone(&self) -> DurationDistrib

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for DurationDistrib

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.