tree_control: Control for strata-finding tree step of 5-STAR algorithm

In rmarceauwest/fiveSTAR: Run 5-STAR Algorithm

Control for strata-finding tree step of 5-STAR algorithm

Description

Parameters for control of conditional inference tree steps of 5-STAR algorithm (3A and 3B) - mainly for passing into ctree and ctree_control functions

Usage

tree_control(minbucket = 40, alpha = c(0.1, 0.2),
  testtype = "Bonferroni", majority = FALSE, maxsurrogate = 3,
  maxdepth = 3, ...)

Arguments

minbucket	Vector of minimum set of weights per terminal node for initial and pruning steps (e.g., minimum number of patients/terminal node for steps 3A and 3B) - if a single number is given, the same value is used for both preliminary and final trees (Default = 40 for both steps)
alpha	vector of significance level for variable selection for tree splits in preliminary and final trees (3A and 3B) - if a single number is given, the same value is used for both preliminary and final trees (Default is (0.1,0.2))
testtype	from ctree_control: "a character specifying how to compute the distribution of the test statistic" (default = "Bonferroni")
majority	ctree control parameter, specifying whether to randomly
maxsurrogate	ctree control parameter defining number of surrogate
maxdepth	Maximum tree depth (default for 5-STAR is 3)
...	additional parameters to be passed into ctree function

Value

A list of control parameters for strata formation step

rmarceauwest/fiveSTAR documentation built on June 30, 2023, 7:38 a.m.