Usage
estimTdiff(x, ...)
"estimTdiff"(x, cov1, cov2, logscale.cov1, logscale.cov2, time0 = 0, conf.level=0.95, ...)
Arguments
x
Object of an appropriate class.
cov1
Vector or matrix with covariates values for which the expectations
of the first survival time are to be computed. It must be a matrix with as many
columns as is the number of covariates (interactions included, intercept excluded)
or the vector of length
equal to the number of covariates (interactions included,
intercept excluded). If matrix is supplied then is assumed that
each row of this matrix gives one covariate combination for the
first survival time.
Intercept is not to be included in cov1. If cov1 is missing an
expectation of a survivor time
for the value of a covariate vector equal to zero is computed. If there is
only intercept in the model, this parameter must be always missing.
cov2
Vector or matrix with covariate values for which the expectations
of the second survival time are to be computed. It must be of same
size as cov1.
logscale.cov1
Vector or matrix with covariate values for the expression of log-scale
(if this depended on covariates) for the first survival time.
It can be omitted in the case that log-scale was common for all observations.
logscale.cov2
Vector or matrix with covariate values for the expression of log-scale
(if this depended on covariates) for the second survival time.
It can be omitted in the case that log-scale was common for all observations.
time0
Starting time of the follow-up as used in the model. I.e. the
model is assumed to be $log(T-time0) = x'beta + sigma*epsilon$
conf.level
confidence level of produced confidence intervals.