rdirichlet {mixstock}R Documentation

Dirichlet deviates

Description

Produces random deviates for the Dirichlet distribution, the multivariate analogue of the beta distribution.

Usage

rdirichlet(n, alpha)

Arguments

n Number of random deviates to generate.
alpha Vector of shape parameters.

Details

The Dirichlet distribution is a multivariate distribution that describes distributions of frequencies. If a multinomial sample n_1, ... n_N is taken from a population, the estimated distribution of frequencies is Dirichlet(n_1+1, ... n_2+N). If {g_i} is a set of Gamma deviates with shape parameters {α_i}, then {g_i}/sum({g_i}) is a Dirichlet deviate.

Value

A matrix of random deviates (each in a distinct row) from rdirichlet

Author(s)

Ben Bolker

Examples

  rdirichlet(5,c(7,4,4))
 

[Package mixstock version 0.9 Index]