Plot an object of class hglasso, hcov, or hbn

Description

This function plots an object hglasso or hcov — graphical representation of the estimated inverse covariance matrix from hglasso, covariance matrix from hcov, or binary network from hbn

Usage

1
2
## S3 method for class 'hglasso'
plot(x, layout=NULL,...) 

Arguments

x

an object of class hglasso, hcov, or hbn.

layout

the layout of the graph to use. If not specified, layout.kamada.kawai is used.

...

additional parameters to be passed to plot.igraph.

Details

This function plots a graphical representation of the estimated inverse covariance matrix or covariance matrix. The hubs are colored in red and has a large vertex size. Features indices for hubs are shown.

Author(s)

Kean Ming Tan

References

Tan et al. (2014). Learning graphical models with hubs. To appear in Journal of Machine Learning Research. arXiv.org/pdf/1402.7349.pdf.

See Also

image.hglasso summary.hglasso hglasso hcov hbn

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
##############################################
# Example from Figure 1 in the manuscript
# A toy example to illustrate the results from 
# Hub Graphical Lasso
##############################################
library(mvtnorm)
set.seed(1)
n=100
p=100

# A network with 4 hubs
Theta<-HubNetwork(p,0.99,4,0.1)$Theta

# Generate data matrix x
x <- rmvnorm(n,rep(0,p),solve(Theta))
x <- scale(x)

# Run Hub Graphical Lasso to estimate the inverse covariance matrix
res1 <- hglasso(cov(x),0.3,0.3,1.5)

# Graphical representation of the estimated Theta
plot(res1,main="conditional independence graph")