Stratified Cox fit on gap = stop - start with stratum = occurrence
number and a common beta: lambda_k(t | H) =
lambda_0k(t - t_(k-1)) exp(beta' X).
Usage
Pwpgt(start, stop, event, X, occurrence, max_iter = 50L, tol = 1e-09)
Arguments
- start, stop
Interval endpoints on the total-time scale.
- event
1 if an event ends the interval, 0 if censored.
- X
Covariate matrix, one row per interval.
- occurrence
Occurrence number of each interval (stratum).
- max_iter, tol
Newton-Raphson controls.
References
Prentice, R. L., Williams, B. J. and Peterson, A. V.
(1981), Biometrika 68(2), 373-379.
Examples
set.seed(1)
occ <- rep(1:3, times = 20)
start <- (occ - 1) * 10
stop <- occ * 10 - runif(60, 0, 3)
event <- rbinom(60, 1, 0.6)
X <- matrix(rnorm(60 * 2), 60, 2)
fit <- Pwpgt(start, stop, event, X, occurrence = occ)
length(fit$estimate) == 2L
#> [1] TRUE
