DirichletProcessVonMises: Create a Dirichlet Mixture of Von Mises distributions

In keesmulder/circbayes: Bayesian Analysis of Circular Data

Description

This is the constructor function to produce a dirichletprocess object with a Von Mises mixture kernel with unknown mean and unknown concentration kappa. The base measure is conjugate to the posterior distribution.

Usage

DirichletProcessVonMises(y, g0Priors = c(0, 0, 1), alphaPriors = c(2,
  4), priorMeanMethod = "integrate", n_samp = 3)

Arguments

y	Data
g0Priors	Base Distribution Priors γ = (μ_0, R_0, n_0)
alphaPriors	Alpha prior parameters. See UpdateAlpha.
priorMeanMethod	Method for marginalization of prior mean. See vonMisesMixtureCreate.
n_samp	Number of Gibbs samples before we assume draws from posterior are i.i.d. See vonMisesMixtureCreate.

Details

The base measure is G_0 (μ, κ \mid γ) = I_0(R κ)^{- n_0} \exp(R_0 κ \cos(μ - μ_0)).

Value

Dirichlet process object

Examples

dp <- DirichletProcessVonMises(rvm(10, 2, 5))
dp

keesmulder/circbayes documentation built on May 30, 2019, 2:04 p.m.