simon2stage: Simon's Two-Stage Design

In lrstat: Power and Sample Size Calculation for Non-Proportional Hazards and Beyond

Simon's Two-Stage Design

Description

Obtains Simon's two-stage minimax, admissible, and optimal designs.

Usage

simon2stage(
  alpha = NA_real_,
  beta = NA_real_,
  piH0 = NA_real_,
  pi = NA_real_,
  n_max = 110L
)

Arguments

alpha	Type I error rate (one-sided).
beta	Type II error rate (1-power).
piH0	Response probability under the null hypothesis.
pi	Response probability under the alternative hypothesis.
n_max	Upper limit for sample size, defaults to 110.

Value

A data frame containing the following variables:

• piH0: Response probability under the null hypothesis.

• pi: Response probability under the alternative hypothesis.

• alpha: The specified one-sided significance level.

• beta: The specified type II error.

• n: Total sample size.

• n1: Stage 1 sample size.

• r1: Futility boundary for stage 1.

• r: Futility boundary for stage 2.

• EN0: Expected sample size under the null hypothesis.

• attainedAlpha: Attained type 1 error.

• power: Attained power.

• PET0: Probability of early stopping under the null hypothesis.

• w_lower: Lower bound of the interval for w.

• w_upper: Upper bound of the interval for w.

• design: Description of the design, e.g., minimax, admissible, or optimal.

Here w is the weight in the objective function: w*n + (1-w)*EN0.

Author(s)

Kaifeng Lu, [e-mail]

Examples

simon2stage(0.05, 0.2, 0.1, 0.3)

lrstat documentation built on Oct. 18, 2024, 9:06 a.m.