binom.nettest: Performes a binomial test with FDR correction for network...
In schw4b/multdyn: Dynamic Graphical Models

Description Usage Arguments Value Examples

Performes a binomial test with FDR correction for network edge occurrence.

1	binom.nettest(adj, alter = "two.sided", fdr = 0.05)

`adj`	adjacency matrix, nodes x nodes x subj, or nodes x nodes x runs x subj.
`alter`	type of binomial test, "two.sided" (default), "less", or "greater"
`fdr`	false discovery rate (FDR) control, default is 0.05.

store list with results.

# Generate some sample binary 5-node network structures for N=20, then perform
# significance testing.
N=20
x = rmdiag(array(rbinom(n=5*5*N, size=1, prob=0.10), dim=c(5,5,N)))
x[1,2,2:N]=1; x[2,3,seq(1,N,2)]=1 # add some consitent edges
A = apply(x, c(1,2), mean)
l = binom.nettest(x)