net.hubs: Get degrees of the most connected nodes of each graph in the...

In fasjem: A Fast and Scalable Joint Estimator for Learning Multiple Related Sparse Gaussian Graphical Models

Description

List the degrees of the hub nodes of each graph in the input list of multiple graphs.

Usage

net.hubs(theta, nhubs = 10)

Arguments

theta	An input list of multiple graphs. Each graph is represented as a pXp matrix. (For example, the result of the fasjem algorithm: a list of pXp matrices in which each matrix represents an estimated sparse inverse covariance matrix.)
nhubs	The number of hubs to be identified of each graph in the input list of multiple graphs.

Value

hubs, a length K list. Each element in this list is a vector of length nhubs whose entries give the degree of the most connected nodes of each graph in the input list of multiple graphs.

Author(s)

Beilun Wang

References

Beilun Wang, Ji Gao, Yanjun Qi (2017). A Fast and Scalable Joint Estimator for Learning Multiple Related Sparse Gaussian Graphical Models. <http://proceedings.mlr.press/v54/wang17e/wang17e.pdf>

Examples

## Not run: 
## load an example multi-task dataset with K=2 tasks, p=100 features, and n=200 samples per task:
data(exampleData)
##run
result = fasjem(X = exampleData, method = "fasjem-g", 0.1, 0.1, 0.1, 0.05, 10)
## get hubs list:
net.hubs(result)

## End(Not run)

fasjem documentation built on May 2, 2019, 9:19 a.m.