rNM: Simulated from Normal Mixture Model

In zrmacc/LFDR: Local False Discovery Rate Estimation

Description

Draws realizations from a normal mixture model. First, the cluster membership is drawn from a multinomial distribution, with mixture proportions specified by pi. Conditional on cluster membership, the observation is drawn from a multivariate normal distribution, with cluster-specific mean and variance. The cluster means are provided using m, and the cluster covariance matrices are provided using v.

Usage

rNM(n, k = 1, pi = NULL, m = 0, v = 1)

Arguments

n	Observations.
k	Number of mixture components. Defaults to 1.
pi	Mixture proportions. If omitted, components are assumed equi-probable.
m	Vector of means, one for each component k.
v	Vector of variances, one for each component k.

Value

Numeric matrix containing the mixture component and the observation.

Examples

## Not run: 
# Two component mixture
Y = rNM(n=1e3,k=2,pi=c(0.95,0.05),m=c(0,2),v=c(1,1));

## End(Not run)

zrmacc/LFDR documentation built on May 3, 2019, 9:01 p.m.