dist

Documentation

Overview

Package dist provides differentiatable distribution models. The package is automatically differentiated by deriv during build.

Index

• Variables
• type Categorical
  • func (dist Categorical) LogZ(alpha []float64) float64
  • func (dist Categorical) Logp(alpha []float64, y float64) float64
  • func (dist Categorical) Logps(alpha []float64, y ...float64) float64
  • func (dist Categorical) Observe(x []float64) float64
• type Dirichlet
  • func (dist Dirichlet) LogZ(alpha []float64) float64
  • func (dist Dirichlet) Logp(alpha []float64, y []float64) float64
  • func (dist Dirichlet) Logps(alpha []float64, y ...[]float64) float64
  • func (dist Dirichlet) Observe(x []float64) float64

Variables

var Bernoulli bernoulli

var Beta beta

Beta distribution, singleton instance

var Binomial binomial

var Cauchy cauchy

Cauchy distribution, singleton instance

var D d

D is a singletone variable of type d. General log-likelihood handling functions are dispatched on d.

var Exponential, Expon exponential

Exponential distribution, singleton instance (Expon is kept for backward compatibility)

var Gamma gamma

Gamma distribution, singleton instance

var Normal normal

Normal distribution, singleton instance

Types

type Categorical

type Categorical struct {
	N int // number of categories
}

Categorical distribution

var Cat Categorical

Categorical distribution, singleton instance; Observe cannot be called on this instance, but Logp and Logps can.

func (Categorical) LogZ

func (dist Categorical) LogZ(alpha []float64) float64

LogZ computes the normalization constant.

func (Categorical) Logp

func (dist Categorical) Logp(
	alpha []float64, y float64,
) float64

Logp computes log pmf of a single observation.

func (Categorical) Logps

func (dist Categorical) Logps(
	alpha []float64, y ...float64,
) float64

Logps computes log pmf of a vector of observations.

func (Categorical) Observe

func (dist Categorical) Observe(x []float64) float64

Observe implements the Model interface

type Dirichlet

type Dirichlet struct {
	N int // number of dimensions
}

Dirichlet distribution

var Dir Dirichlet

Dirichlet distribution, singleton instance; Observe cannot be called on this instance, but Logp and Logps can.

func (Dirichlet) LogZ

func (dist Dirichlet) LogZ(alpha []float64) float64

LogZ computes the normalization constant.

func (Dirichlet) Logp

func (dist Dirichlet) Logp(alpha []float64, y []float64) float64

Logp computes log pdf of a single observation.

func (Dirichlet) Logps

func (dist Dirichlet) Logps(alpha []float64, y ...[]float64) float64

Logps computes log pdf of a vector of observations.

func (Dirichlet) Observe

func (dist Dirichlet) Observe(x []float64) float64

Observe implements the Model interface. The parameters are alpha and observations, flattened.