joineRML (version 0.1.1)

bootSE: Standard errors via bootstrap for an mjoint object

Description

This function takes a model fit from an mjoint object and calculates standard errors and confidence intervals for the main longitudinal and survival coefficient parameters, including the latent association parameters, using bootstrapping (Efron and Tibshirani, 2000).

Usage

bootSE(object, nboot = 100, ci = 0.95, use.mle = TRUE, verbose = FALSE, control = list(), progress = TRUE, ...)

Arguments

object

an object inheriting from class mjoint for a joint model of time-to-event and multivariate longitudinal data.

nboot

the number of bootstrap samples. Default is nboot=100.

ci

the confidence interval to be estimated using the percentile-method. Default is ci=0.95 for a 95% confidence interval.

use.mle

logical: should the algorithm use the maximiser from the converged model in object as initial values for coefficients in each bootstrap iteration. Default is use.mle=TRUE.

verbose

logical: if TRUE, the parameter estimates and other convergence statistics are value are printed at each iteration of the MCEM algorithm. Default is FALSE.

control

a list of control values with components:

progress

logical: should a progress bar be shown on the console to indicate the percentage of bootstrap iterations completed? Default is progress=TRUE.

...

options passed to the control argument.

Value

An object of class bootSE.

Details

Standard errors and confidence intervals are obtained by repeated fitting of the requisite joint model to bootstrap samples of the original longitudinal and time-to-event data. Note that bootstrap is done by sampling subjects, not individual records.

References

Efron B, Tibshirani R. An Introduction to the Bootstrap. 2000; Boca Raton, FL: Chapman & Hall/CRC.

See Also

mjoint for approximate standard errors.

Examples

## Not run: 
# # Fit a joint model with bivariate longitudinal outcomes
# 
# data(heart.valve)
# hvd <- heart.valve[!is.na(heart.valve$log.grad) & !is.na(heart.valve$log.lvmi), ]
# 
# fit2 <- mjoint(
#     formLongFixed = list("grad" = log.grad ~ time + sex + hs,
#                          "lvmi" = log.lvmi ~ time + sex),
#     formLongRandom = list("grad" = ~ 1 | num,
#                           "lvmi" = ~ time | num),
#     formSurv = Surv(fuyrs, status) ~ age,
#     data = list(hvd, hvd),
#     inits = list("gamma" = c(0.11, 1.51, 0.80)),
#     timeVar = "time",
#     verbose = TRUE)
# 
# fit2.boot <- bootSE(fit1, 50, use.mle = TRUE, control = list(
#     earlyPhase = 25, convCrit = "either",
#     tol0 = 6e-03, tol2 = 6e-03, mcmaxIter = 60))
# ## End(Not run)