ntbft.boot: Net benefit Bootstrap method to correct overfitting

In ProfessionalFarmer/loonR: Common function in bioinformatics analysis

Net benefit Bootstrap method to correct overfitting

Description

Net benefit Bootstrap method to correct overfitting

Usage

ntbft.boot(
  outcome,
  pred,
  boots = 500,
  xstart = 0.01,
  xstop = 0.99,
  step = 0.01,
  type = "treated",
  model.name = "Model"
)

Arguments

boots	500
xstart	the starting point of the threshold probability, the default value is 0.01.
xstop	the end point of the threshold probability, the default value is 0.99
step	a numerical value specifying the incremental step of the threshold probability, the default value is 0.01
type	controls the type of net benefit to be computed. The allowed values correspond to the treated (“treated”), untreated (“untreated”) and overall (“overall”) patients, or to the ADAPT index (“adapt”). The default is the “treated”
model.name
response	Should be 0 and 1

Examples

data(LIRI)

pred <- as.vector( unlist( LIRI[,c(3)] ) )
ntbft.boot(LIRI$status, pred)

ProfessionalFarmer/loonR documentation built on Oct. 9, 2024, 9:56 p.m.