rmdm: Generate random samples from a mixture of Dirchelet...

In dwinter/mdm: Mixtures of Dirichelet-Multinomial Distributions

Description

Generate random samples from a mixture of Dirchelet Multinomials

Usage

rmdm(n, m, f, phi = NULL, p = NULL, scale = NULL)

Arguments

n	number of observations
m	observation sizes vector possibly of length one.
f	mixture proportions
phi	Overdispersion paramater. Value must be in [0,1], where phi=0 is 'pure' Multinomial distribuion and increasing values of phi lead to increasingly over-dispersed distributions.
p	matrix or vector of proportions
scale	Scale parameters, vector of matrix.

Value

A matrix of n-rows, each row representing a random draw from

Examples

#simulate reads with 1% sequencing error
#from AT-biased genome
genotypes <- matrix( 0.01/3, nrow=4, ncol=4)
diag(genotypes) <- 0.99
nfreq <- c(0.4, 0.1, 0.1, 0.4)
rmdm(10, 40, phi=0,    f=nfreq, p=genotypes)
rmdm(10, 40, phi=0.01, f=nfreq, p=genotypes)

dwinter/mdm documentation built on May 15, 2019, 4:54 p.m.