prune: Function to Prune Using the Score Statistic

In TimeVTree: Survival Analysis of Time Varying Coefficients Using a Tree-Based Approach

Description

This function merges over-segmented intervals to create optimally pruned subtrees.

Usage

prune(fulltree)

Arguments

fulltree

output from output.coxphout

Details

prune uses the CART algorithm and -log (partial likelihood) as cost to find the optimally pruned subtrees.

Value

prune returns a matrix with the following columns, where each row is an optimally pruned subtree:

K	subtrees number 1, 2, etc. Tree #1 is the full tree
N[1]	Number of terminal nodes
alpha	penalty parameter corresponding to the subtree
S[1]	-log(partial likelihood) of the subtree
pruneoff	Node that was removed from the previous larger subtree to obtain the current subtree

References

Xu, R. and Adak, S. (2002), Survival Analysis with Time-Varying Regression Effects Using a Tree-Based Approach. Biometrics, 58: 305-315.

Examples

##Call in alcohol data set
data('alcohol')
require(survival)

coxtree <- coxph.tree(alcohol[,'time'], alcohol[,'event'], 
                      x = alcohol[,'alc', drop = FALSE], D = 4)
nodetree <- output.coxphout(coxtree)

subtrees <- prune(nodetree)

TimeVTree documentation built on May 2, 2019, 2:17 a.m.