survBayes {survBayes} | R Documentation |
Fits a proportional hazards model to time to event data by a Bayesian approach. Right and interval censored data and a lognormal frailty term can be fitted.
survBayes(formula = formula(data), data = parent.frame(), burn.in = 1000, number.sample = 1000, max.grid.size = 20, control, control.frailty, seed.set = 100, ...)
formula |
a formula object, with the response on the left of a ~ operator, and the terms on the right.
The response must be a survival object of type "right" or "interval"
as returned by the Surv function. |
data |
a data.frame in which to interpret the variables named in the formula |
burn.in |
burn.in |
number.sample |
number of sample |
max.grid.size |
number of grid points |
control |
Object of class control specifying iteration limit and other control options. Default is survBayes.control(...). |
control.frailty |
Object of class control.frailty specifying parameters for the priors of frailties and other control options.
Default is survBayes.control.lognormal.frailty(...) or survBayes.control.gamma.frailty(...). |
seed.set |
setting of the seed of the random number generator |
... |
further parameters |
Fits a proportional hazards model to time to event data by a Bayesian approach.
The time axis is split into max.grid.size
intervals and the
log baseline hazard is assumed to be cubic spline penalized by an auto regressive process of order one.
Right and interval censored data and a lognormal or gamma frailty term can be fitted.
In case of interval censored data the assumed observation times are
augmented by a piecewise exponential distribution conditioned on the respective interval.
The returned values are, if appropriate
t.where |
used grid points |
beta |
samples of the vector of covariates |
lbh.coef |
samples of the log baseline hazard coefficients at the grid points |
sigma.lbh |
samples of sigma.lbh.0 and sigma.lbh.1 |
alpha.cluster |
samples of the frailty values |
sigma.cluster |
samples of frailty variance |
z.cluster |
samples of the frailty values |
mu.cluster |
samples of the rate and shape of the gamma prior |
m.h.performance |
The performance of the Metropolis-Hasting steps is checked for beta, lbh and, if appropriate, alpha.cluster or mu.cluster |
V. Henschel, Ch. Heiss, U. Mansmann
data(aneurism.data) control<-survBayes.control(sigma.lbh.1=0.01,rate.sigma.lbh.1 = 1e-3, shape.sigma.lbh.1 = 1e-3) aneurism.res<-survBayes(Surv(left,right,cens*3,type="interval")~mo+loc+frailty(gr,dist="gamma"),data=aneurism.data,burn.in=0,number.sample=10,control=control)