mv_gmm_gen: Generates data from a multi-view Gaussian mixture model

In multiviewtest: Hypothesis Tests for Association Between Subgroups in Two Data Views

Description

Generates data from a multi-view Gaussian mixture model with n observations and two views.

Usage

mv_gmm_gen(n, Pi, mu1, mu2, Sigma1, Sigma2)

Arguments

n	number of observations
Pi	K1 x K2 matrix where the (k, k')th entry contains the probability of an observation belonging to cluster k in View 1 and cluster k' in View 2
mu1	p1 x K1 matrix where the columns contain the K1 cluster means in View 1
mu2	p2 x K2 matrix where the columns contain the K2 cluster means in View 2
Sigma1	p1 x p1 matrix containing the covariance matrix for View 1
Sigma2	p2 x p2 matrix containing the covariance matrix for View 2

Value

A list containing the following components:

data	A list with two items: the view 1 n x p1 multivariate data set and the view 2 n x p2 multivariate data set
clusters	A list with two items: the view 1 cluster memberships and the view 2 cluster memberships

References

Gao, L.L., Bien, J., Witten, D. (2019) Are Clusterings of Multiple Data Views Independent? Biostatistics, <DOI:10.1093/biostatistics/kxz001>

Gao, L.L., Witten, D., Bien, J. Testing for Association in Multi-View Network Data, preprint.

Examples

# 25 draws from a two-view Gaussian mixture model where the clusters are independent
n <- 25
Pi <-  tcrossprod(c(0.5, 0.5), c(0.25, 0.25, 0.5))
mu1 <- cbind(c(2, 2), c(-2, 2))
mu2 <- cbind(c(0, 1), c(1, 0), c(-1, 0))
Sigma1 <- diag(rep(1, 2))
Sigma2 <- diag(rep(0.5, 2))

mv_gmm_gen(n, Pi, mu1, mu2, Sigma1, Sigma2)

multiviewtest documentation built on Oct. 13, 2021, 5:08 p.m.