ExpFam_dist_canonical: Exponential Family distribution in canonical form

Description Usage Arguments

Description

The class defines the distribution in the canonical representation, with the density function defined as: d(x|η) = h(x) exp{(η^T S(x) - A(η))}

Usage

1
2
ExpFam_dist_canonical(h, A.from.eta, S, eta.dim, A.grad, A.hess, eta.in.domain,
  family.name)

Arguments

h

(function) the measure

A.from.eta

(function) value of log-normalizer for natural parameter in canonical parametrization

S

(function) vector of sufficient statistics

eta.dim

(numeric), number of parameters of the exponential family being defined

A.grad

(function) gradient of the function A.from.eta

A.hess

(function) hessian of the function A.from.eta

eta.in.domain

(function) characteristic function of η

family.name

(character) the family name (e.g. Bernoulli)


cezden/ExpFamilyLab documentation built on May 13, 2019, 3:07 p.m.