Struct Sampler 

pub struct Sampler { /* private fields */ }

Thread-local statistical distribution sampler backed by a high-quality PRNG.

Implementations

impl Sampler

pub fn new() -> Self

Creates a new sampler instance with a thread-local pseudorandom number generator.

pub fn sample_standard_normal(&mut self) -> f64

Generates a single sample from the standard normal distribution N(0, 1).

pub fn gamma(&mut self, shape: f64, scale: f64) -> f64

Gamma(shape, scale). Preconditions: shape > 0, scale > 0.

pub fn chi2(&mut self, df: f64) -> f64

Chi-square(df) == Gamma(df/2, 2). Preconditions: df > 0.

pub fn standard_normal_vec(&mut self, dim: usize) -> Vec64<f64>

Vector of iid N(0,1) samples of length dim.

pub fn dirichlet(&mut self, alpha: &[f64]) -> Vec64<f64>

Dirichlet-distributed probability vector via normalised gamma sampling.

Preconditions: alpha non-empty; all entries finite and > 0.