bsDataStep: Generate bootstrap distribution of model parameters to fitted...

In incubate: Parametric Time-to-Event Analysis with Variable Incubation Phases

Generate bootstrap distribution of model parameters to fitted incubate model.

Description

Bootstrap data are here estimated coefficients from models fitted to bootstrap samples. The bootstrap data is used to make bootstrap inference in the second step. It is an internal function, the main entry point is confint.incubate_fit().

Usage

bsDataStep(
  object,
  bs_data = c("parametric", "ordinary"),
  R,
  useBoot = FALSE,
  smd_factor = 0.25
)

Arguments

object	an incubate_fit-object
bs_data	character. Which type of bootstrap method to generate data?
R	integer. Number of bootstrapped model coefficient estimates
useBoot	flag. Do you want to use the boot-package? Default value is FALSE.
smd_factor	numeric. smooth-delay factor: influence the amount of smoothing. 0 means no smoothing at all. Default is 0.25 (as was optimal in simulation for log-quantile together with log-delay-shift = 5)

Value

bootstrap data, either as matrix or of class boot (depending on the useBoot-flag)

incubate documentation built on Aug. 21, 2023, 5:11 p.m.