Struct rv::dist::Categorical

pub struct Categorical {
    pub ln_weights: Vec<f64>,
}

Distribution over unordered values in [0, k)

Fields

ln_weights: Vec<f64>

Methods

impl Categorical

pub fn new(weights: &[f64]) -> Result<Self>

Construct a new Categorical distribution from weights

Examples

extern crate assert;
extern crate rv;

use rv::traits::*;
use rv::dist::Categorical;

let weights: Vec<f64> = vec![4.0, 2.0, 3.0, 1.0];
let cat = Categorical::new(&weights).unwrap();

assert!(cat.contains(&0_u8));
assert!(cat.contains(&3_u8));
assert!(!cat.contains(&4_u8));

assert::close(cat.pmf(&0_u8), 0.4, 10E-12);

pub fn from_ln_weights(ln_weights: Vec<f64>) -> Result<Self>

Build a Categorical distribution from normalized log weights

pub fn uniform(k: usize) -> Self

Creates a Categorical distribution over [0, k) with uniform weights

pub fn weights(&self) -> Vec<f64>

Return the weights (exp(ln_weights))

pub fn k(&self) -> usize

Trait Implementations

impl Debug for Categorical

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

Formats the value using the given formatter.

impl Clone for Categorical

fn clone(&self) -> Categorical

Returns a copy of the value.

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

Performs copy-assignment from source.

impl<X: CategoricalDatum> Rv<X> for Categorical

fn ln_f(&self, x: &X) -> f64

Un-normalized probability function

fn ln_normalizer(&self) -> f64

The log of the constant term in the PDF/PMF. Should not be a function of any of the parameters.

fn draw<R: Rng>(&self, rng: &mut R) -> X

Single draw from the Rv

fn sample<R: Rng>(&self, n: usize, rng: &mut R) -> Vec<X>

Multiple draws of the Rv

fn f(&self, x: &X) -> f64

Un-normalized probability function

fn normalizer(&self) -> f64

The constant term in the PDF/PMF. Should not be a function of any of the parameters.

impl<X: CategoricalDatum> Support<X> for Categorical

fn contains(&self, x: &X) -> bool

Returns true if x is in the support of the Rv

impl<X: CategoricalDatum> DiscreteDistr<X> for Categorical

fn pmf(&self, x: &X) -> f64

Probability mass function (PMF) at x

fn ln_pmf(&self, x: &X) -> f64

Natural logarithm of the probability mass function (PMF)

impl<X: CategoricalDatum> Mode<X> for Categorical

fn mode(&self) -> Option<X>

Returns None if the mode is undefined or is not a single value

impl Entropy for Categorical

fn entropy(&self) -> f64

The entropy, H(X)

impl<X: CategoricalDatum> HasSuffStat<X> for Categorical

type Stat = CategoricalSuffStat

fn empty_suffstat(&self) -> Self::Stat

impl Rv<Categorical> for SymmetricDirichlet

fn ln_f(&self, x: &Categorical) -> f64

Un-normalized probability function

fn ln_normalizer(&self) -> f64

The log of the constant term in the PDF/PMF. Should not be a function of any of the parameters.

fn draw<R: Rng>(&self, rng: &mut R) -> Categorical

Single draw from the Rv

fn f(&self, x: &X) -> f64

Un-normalized probability function

fn normalizer(&self) -> f64

The constant term in the PDF/PMF. Should not be a function of any of the parameters.

fn sample<R: Rng>(&self, n: usize, rng: &mut R) -> Vec<X>

Multiple draws of the Rv

impl<X: CategoricalDatum> ConjugatePrior<X, Categorical> for SymmetricDirichlet

type Posterior = Dirichlet

fn posterior(&self, x: &CategoricalData<X>) -> Self::Posterior

Computes the posterior distribution from the data

fn ln_m(&self, x: &CategoricalData<X>) -> f64

Log marginal likelihood

fn ln_pp(&self, y: &X, x: &CategoricalData<X>) -> f64

Log posterior predictive of y given x

fn m(&self, x: &DataOrSuffStat<X, Fx>) -> f64

Marginal likelihood of x

fn pp(&self, y: &X, x: &DataOrSuffStat<X, Fx>) -> f64

Posterior Predictive distribution

impl Rv<Categorical> for Dirichlet

fn ln_f(&self, x: &Categorical) -> f64

Un-normalized probability function

fn ln_normalizer(&self) -> f64

The log of the constant term in the PDF/PMF. Should not be a function of any of the parameters.

fn draw<R: Rng>(&self, rng: &mut R) -> Categorical

Single draw from the Rv

fn f(&self, x: &X) -> f64

Un-normalized probability function

fn normalizer(&self) -> f64

The constant term in the PDF/PMF. Should not be a function of any of the parameters.

fn sample<R: Rng>(&self, n: usize, rng: &mut R) -> Vec<X>

Multiple draws of the Rv

impl<X: CategoricalDatum> ConjugatePrior<X, Categorical> for Dirichlet

type Posterior = Self

fn posterior(&self, x: &CategoricalData<X>) -> Self::Posterior

Computes the posterior distribution from the data

fn ln_m(&self, x: &CategoricalData<X>) -> f64

Log marginal likelihood

fn ln_pp(&self, y: &X, x: &CategoricalData<X>) -> f64

Log posterior predictive of y given x

fn m(&self, x: &DataOrSuffStat<X, Fx>) -> f64

Marginal likelihood of x

fn pp(&self, y: &X, x: &DataOrSuffStat<X, Fx>) -> f64

Posterior Predictive distribution