GeneNet (version 1.2.16)

ggm.simulate.pcor: Graphical Gaussian Models: Simulation of Networks

Description

ggm.simulate.pcor generates a random matrix of partial correlations that corresponds to a GGM network of a given size (num.nodes) with a specified fraction of non-zero edges. The diagonal entries of the output matrix contain 1.

If stdprec=TRUE then the standardised precision matrix is returned instead of the patrix of partial correlations.

Usage

ggm.simulate.pcor(num.nodes, etaA=0.05, stdprec=FALSE)

Arguments

num.nodes

number of nodes in the network

etaA

fraction of edges with non-zero partial correlation (default: 0.05)

stdprec

return standardised precision matrix, rather than matrix of partial correlations

Value

A positive partial correlation matrix (diagonal 1) with positive definite underlying precision matrix.

If stdprec=TRUE then the standardised precision matrix is returned instead.

Details

The simulation of the partial correlation matrix works by generating a diagonally dominant matrix as a positive definite precision matrix (inverse covariance matrix), which is subsequently standardized and transformed into the matrix of partial correlations. For the full algorithm see Sch\"afer and Strimmer (2005).

References

Sch\"afer, J., and Strimmer, K. (2005). An empirical Bayes approach to inferring large-scale gene association networks. Bioinformatics 21:754-764.

See Also

ggm.simulate.data,ggm.estimate.pcor.

Examples

# NOT RUN {
# load GeneNet library
library("GeneNet")

# generate random network with 40 nodes 
# it contains 780=40*39/2 edges of which 5 percent (=39) are non-zero
true.pcor <- ggm.simulate.pcor(40)
  
# simulate data set with 40 observations
m.sim <- ggm.simulate.data(40, true.pcor)

# simple estimate of partial correlations
estimated.pcor <- cor2pcor( cor(m.sim) )

# comparison of estimated and true values
sum((true.pcor-estimated.pcor)^2)

# a slightly better estimate ...
estimated.pcor.2 <- ggm.estimate.pcor(m.sim)
sum((true.pcor-estimated.pcor.2)^2)

# }