Per-occurrence marginal Cox models on the total-time scale plus the
common-effect summary: the stratified fit with a single beta across
occurrence strata (WLW Section 3).
Usage
Wlwmm(time, event, X, occurrence, max_iter = 50L, tol = 1e-09)
Arguments
- time
Total time on study for each row.
- event
1 if the row's occurrence was observed, else 0.
- X
Covariate matrix, one row per (subject, occurrence).
- occurrence
Occurrence number of each row.
- max_iter, tol
Newton-Raphson controls.
Value
List with estimate (common beta), se,
cov, per_event_beta (named list), loglik,
n_iter, n_events, method.
References
Wei, L. J., Lin, D. Y. and Weissfeld, L. (1989), JASA
84(408), 1065-1073.
Examples
set.seed(1)
occ <- rep(1:3, times = 20)
stop <- occ * 10 - runif(60, 0, 3)
event <- rbinom(60, 1, 0.6)
X <- matrix(rnorm(60 * 2), 60, 2)
fit <- Wlwmm(time = stop, event, X, occurrence = occ)
is.list(fit)
#> [1] TRUE
