R/max_splitter.R

In zsiders/EnsembleRandomForests: Ensemble Random Forests

#' Determines the maximum sample size in the second RF bag
#' 
#' @description Takes a data.frame object determines the maximum split
#' 
#' @param v A data frame object returned from erf_data_prep()
#' @param p A numeric value between (0,1), default is just below 90%
#' @param nmax A numeric value > 0 specifying the maximum number of observations per bag, default is 1e4
#' 
#' @return A numeric value specifying the maximum split
#' @export
#' 
#' @examples
#' data <- erf_data_prep(df=simData$samples, var='obs', covariates=grep('cov', colnames(simData$samples), value=TRUE))
#' max_splitter(data)
#' max_splitter(data, p=0.6)
#' 
max_splitter <- function(v, p=0.89, nmax=1e4){
	t <- table(v[,1]) #assumes variable of interest is in the first column
	#p is raised to the power of 2 for the two bagging events
	max_split <- pmin(nmax,as.numeric(floor(min(t*p^2))))
	return(max_split)
}