power.weighted.nabla: Exact weighted power; sensitivity parameterization

In cpalmes/schuirmann.constant: Asymptotic Properties of the TOST and the Schuirmann Constant

Description

Calculates the exact power of an unpaired, balanced TOST with n samples in each group, where the true mean-difference is modeled with an a-priori density rather than a fixed mean-difference. The calculation is performed in the TOST sensitivity parameterization.

Usage

power.weighted.nabla(alpha, density, nabla, n)

Arguments

alpha	type I error
density	a-priori density for the mean-difference
nabla	TOST sensitivity index
n	sample size in each group

Value

Exact power of the TOST for a weighted mean-difference

Examples

rho <- apriori.density('U')
power.weighted.nabla(alpha = 0.05, density = rho, nabla = 8, n = 22)
## Since we are performing a weighted power calculation, the r respectively 
## theta0 value is not needed. Thus, the following calculation does not depend
## on r.
l <- nabla.to.standard(r = 0.99, nabla = 8, theta1 = -1, theta2 = 4, n = 22)
power.weighted(alpha = 0.05, theta1 = -1, theta2 = 4, density = rho, sigma = l$sigma, n = 22)

cpalmes/schuirmann.constant documentation built on Dec. 31, 2020, 10:07 p.m.