tronco.kfold.posterr {TRONCO}R Documentation

tronco.kfold.posterr. For details and examples regarding the statistical assesment of an inferred model, we refer to the Vignette Section 7.

Description

Perform a k-fold cross-validation using the function bn.cv and scan every node to estimate its posterior classification error.

Usage

tronco.kfold.posterr(x, models = names(as.models(x)),
  events = as.events(x), runs = 10, k = 10, cores.ratio = 1,
  silent = FALSE)

Arguments

x

A reconstructed model (the output of tronco.capri)

models

The names of the selected regularizers (bic, aic or caprese)

events

a list of event

runs

a positive integer number, the number of times cross-validation will be run

k

a positive integer number, the number of groups into which the data will be split

cores.ratio

Percentage of cores to use. coresRate * (numCores - 1)

silent

A parameter to disable/enable verbose messages.

Examples

data(test_model)
tronco.kfold.posterr(test_model, k = 2, runs = 2)


[Package TRONCO version 2.16.2 Index]