Struct HistogramDistribution

pub struct HistogramDistribution { /* private fields */ }

Distribution described by non-normalised histogram

The distribution is given as a table of x and pdf which follows histogram interpolation. For x ∈ [xᵢ₊₁; xᵢ] probability if pdf(x) = pdfᵢ. The cumulative distribution function becomes piece-wise linear which makes sampling form the table quite easy.

To support better approximation of different pdfs, the grid is not equal-spaced in general. Hence binary search is needed to find correct bin.

Implementations

impl HistogramDistribution

pub fn new(x: Vec<f64>, pdf: Vec<f64>) -> Result<Self, SetupError>

Create a new instance of the histogram distribution

The cumulative distribution function will be calculated.

Condition x.len() == pdf.len() > 1, must be met. Thus the last value in the pdf vector will be ignored.

pub fn sample_with_value<RNG>(&self, rng: &mut RNG) -> (f64, f64)

where RNG: Rng + ?Sized,

Sample a value from the histogram and return the value of non-normalised probability

Result

Tuple (s, p) where s is the sample and p probability in the bin

We need a way to sample while returning probability value in the bin as well to implement rejection sampling scheme without repeating a binary search of the grid.

Trait Implementations

impl Clone for HistogramDistribution

fn clone(&self) -> HistogramDistribution

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for HistogramDistribution

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

Formats the value using the given formatter.

impl Distribution<f64> for HistogramDistribution

Draws samples from the histogram distribution

fn sample<R: Rng + ?Sized>(&self, rng: &mut R) -> f64

Generate a random value of T, using rng as the source of randomness.

fn sample_iter<R>(self, rng: R) -> DistIter<Self, R, T>

where R: Rng, Self: Sized,

Create an iterator that generates random values of T, using rng as the source of randomness.

fn map<F, S>(self, func: F) -> DistMap<Self, F, T, S>

where F: Fn(T) -> S, Self: Sized,

Create a distribution of values of ‘S’ by mapping the output of Self through the closure F