Check if a design is consonant — check

For some CER design, see if the defined test sequence is consonant, either from just it's graph and weights, also using the correlation structure, or after some adaption.

Usage

check_consonance(
  design,
  adapted = design$adaptions,
  stage = c("both", "interim", "final"),
  use_weights = FALSE
)

Arguments

design: A cer_design object
adapted: Boolean, use the adapted or original values
stage: For which stage should the consonance be tested? Can be one of "both", "interim", "final". No effect if weights = TRUE.
use_weights: Boolean, if TRUE uses the weights instead of the actual bounds. This is the same when no correlation structure is used, and else checks consonance of the counterfactual where the graph used is the same, but no use of correlation structure is made

Value

A boolean value, TRUE if the design (for the specified weights) is consonant, else FALSE

Details

For every combination of hypothesis of a design, a intersection hypothesis is being built that is keeping the specified alpha level, and only when all intersection hypothesis holding a specific hypothesis are rejected, that hypothesis is rejected. Consonance now means that whenever a intersection hypothesis is rejected, at least one intersection hypothesis that is a superset of this one is rejected. If a design is not consonant, it is in some sense not optimal

Examples

as <- function(x,t) 2-2*pnorm(qnorm(1-x/2)/sqrt(t))
design <- cer_design(
 correlation=rbind(H1=c(1, NA),
                   H2=c(NA, 1)),
 weights=c(2/3, 1/3),
 alpha=0.05,
 test_m=rbind(c(0, 1),
              c(1, 0)),
 alpha_spending_f=as,
 t=0.5)

check_consonance(design)
#> [1] TRUE
check_consonance(design, stage = "final")
#> [1] TRUE