OneWaySurvfitObject {TwoWaySurvival} | R Documentation |
An object of class 'OneWaySurvfit' returned as a list by the 'OneWaySurvfitCreate' function and representing a fitted one-way hazard model with varying coefficients. It has methods for generic functions print, plot and summary.
fix.coef |
named vector of not penalized parameters of the model. It has the value 'NULL' if B-spline bases are chosen for modelling. |
random.coef |
named vector of penalized parameters of the model. |
penalty |
named vector of penalty values for the random parts of the model. |
var.fix |
estimated variances of not penalized parameters of the model. It has the value 'NULL' if B-spline bases are chosen for modelling. |
var.random |
estimated variances of the predicted values of (penalized) random components of the model. |
log.lik.margin.start |
initial value of the marginal log-likelihood. |
log.lik.margin |
the value of the marginal log-likelihood at optimum. |
df |
estimated degrees of freedom for the parts of the model. |
df.total |
total or summed estimated degrees of freedom. |
niter.epoch |
number of outer loops in optimization routine. |
varying.frame |
data frame with named columns, according to the names of the varying coefficients. It is needed for plotting of the smooth components. |
deviation.frame |
data frame with named columns, according to the names of the varying coefficients. It is needed for plotting of the confidence bands of the smooth components. |
grid.frame |
data frame with just one column, for survival time, needed for plotting issues. |
p |
number of covariates; for factors including their categories (excluding reference category). |
factor.names |
covariate names; for factors the names of categories (excluding reference category). |
Pavel Khomski <pkhomski@wiwi.uni-bielefeld.de>
Kauerman G. (2005). Penalised Spline Fitting in Multivariable Survival Models with Varying Coefficients Computational Statistics and Data Analysis, 49, 169-186.
OneWaySurv
, OneWaySurvfitCreate