Calculate bounds for preplanned test — cer_prep

Calculates the bounds and cJ values for the p-values for the interim test and the planned final test

Usage

cer_prep_bounds(correlation, weights, alpha, t)

Arguments

correlation: matrix describing the correlation structure of the hypotheses
weights: list of weights of the given hypotheses, same length as hypotheses
alpha: list of length 2 describing the amount of alpha spent that the test of the hypotheses should adhere to in the first and second stage respectively
t: information fraction at which the first stage test is planned

Value

A list with the following elements:

bounds_1: vector of same length as weights with bounds for the first interim test
cJ1: number that gets multiplied by the weights to get bounds_1
bounds_1: vector of same lenght as weights with bounds for the planned final test
cJ2: number that gets multiplied by the weights to get bounds_2

Examples

#simple non-parametric bonferroni
cer_prep_bounds(
    correlation = rbind(c(1,0), c(0,1)),
    weights = c(0.5,0.5),
    alpha = c(0.001525323, 0.025),
    t = 0.5)
#> $bounds_1
#> [1] 0.000762953 0.000762953
#> 
#> $bounds_2
#> [1] 0.0122729 0.0122729
#> 
#> $cJ1
#> [1] 0.001525906
#> 
#> $cJ2
#> [1] 0.0245458
#> 

#weighted bonferroni with correlation 0.5
cer_prep_bounds(
    correlation = rbind(c(1,0.5), c(0.5,1)),
    weights = c(2/3,1/3),
    alpha = c(0.001525323, 0.025),
    t = 0.5)
#> $bounds_1
#> [1] 0.0010404644 0.0005202322
#> 
#> $bounds_2
#> [1] 0.017430042 0.008715021
#> 
#> $cJ1
#> [1] 0.001560697
#> 
#> $cJ2
#> [1] 0.02614506
#>