Patterns (version 1.2)

cutoff: Choose the best cutoff

Description

Allows estimating the best cutoff. For a sequence of cutoff, the p value corresponding to each cutoff value of the sequence. Mainly recommended for single time cascade networks. To achieve more sparsity in other settings, please use a fiiting function based on the stability selection or selectboost algorithms.

Usage

cutoff(Omega,...)

Arguments

Omega

a network object

...

Optional arguments:

sequence

a vector corresponding to the sequence of cutoffs that will be tested.

x_min

an integer ; only values over x_min are further retained for performing the test.

Value

A list containing two objects :

p.value

the p values corresponding to the sequence of cutoff

p.value.inter

the smoothed p value vector, using the loess function

Examples

# NOT RUN {
		data(network)
		cutoff(network)
		#See vignette for more details
# }