In keaven/gsDesign2: Group Sequential Design with Non-Constant Effect

library(tibble)
library(dplyr)
#library(gsDesign2)
devtools::load_all()

Introduction of gs_b()

gs_b() can be used to derive fixed boundary in fixed/group sequential design. It is usually used in the upper = ... and lower = ... arguments in + gs_power_npe() + gs_design_npe() + gs_power_ahr() + gs_design_ahr() + gs_power_wlr() + gs_design_wlr() + gs_power_combo() + gs_design_combo()

Usage of gs_b()

Example 1

Assume it is a group sequential design with 3 analysis, one can input its upper bound as a vector c(4, 3, 2) by using gs_b() as follows.

gs_b(par = 4:2)

Example 2

In the above example, one can assign the upper bound at the second analysis by

gs_b(par = 4:2, k = 2)

Example 3

Generate an efficacy bound using a spending function. Use Lan-DeMets spending approximation of O'Brien-Fleming bound as 50\%, 75\% and 100\% of final spending

# information fraction
IF <- c(.5, .75, 1)
# Lan-DeMets spending approximation of O'Brien-Fleming
par <- gsDesign::gsDesign(alpha = .025, k = length(IF), 
                          test.type = 1, sfu = gsDesign::sfLDOF, 
                          timing = IF)$upper$bound
gs_b(par = par)

Inner Logic of gs_b()

gs_b is a short function with 2 key arguments: par = ... and k = ...

if(is.null(k)){
  return(par)
}else{ 
  return(par[k])
}

keaven/gsDesign2 documentation built on Oct. 13, 2022, 8:42 p.m.