seqp: Compute Boundary Crossing Probabilities

View source: R/seqp.R

seqpR Documentation

Compute Boundary Crossing Probabilities

Description

Computes probabilities of crossing arbitrary user specified upper and/or lower boundaries

Usage

seqp(inf.times, upper, lower = NULL, eta = 0)

Arguments

inf.times

Information times of analyses

upper

Upper boundary (for rejecting H0) on standard normal scale

lower

Lower boundary (for rejecting HA) on standard normal scale

eta

The mean parameter on the Brownian motion process scale.

Details

For a group sequential design with arbitrary boundaries, computes the probability of crossing the upper and lower boundary at each analysis. The probability at each analysis is the probability of not crossing either boundary at earlier analyses and crossing at the current analysis.

eta should be 0 to compute probabilities under the null, and should be set to a value which is a function of the alternative on a transformed scale to compute probabilities under the alternative. For survival studies, eta can be computed using the functions seqopr or lr.inf. (eta is the mean of Y(1), where Y(t) is the partial sum process standardized to have variance 1 at t=1.)

Value

A length(inf.times) by 5 matrix with columns giving the information times, the upper boundary, the lower boundary, the probabilities of crossing the upper boundary, and the probabilities of crossing the lower boundary.

See Also

sequse; seqopr; lr.inf

Examples

seqp((1:4) / 4, c(2, 2, 3, 2), eta = 3)
seqp((1:4) / 4, c(2, 2, 3, 2), eta = 0)


raredd/desmon documentation built on May 7, 2024, 3:46 p.m.