Survival probability conditional to the observed data estimation for correlated right censored data.
Csurv(
Y,
M,
censor,
group = NULL,
t = 0,
w,
knots = 10,
kappa = 10000,
method = "marg",
RandDist = "Gamma",
hazard = "Splines",
maxit = 300
)Return vector of estimated event status and its complement.
a numeric vector of event-times or observed times.
a numeric vector of (bio)marker or risk score values.
a vector of censoring indicator, 1 if event, 0 otherwise.
a categorical vector of group/cluster.
a scalar time for prediction. The default value is 0.
a scalar window for prediction.
a scalar for specifying the number of knots to use. Value required in the penalized likelihood estimation. It corresponds to the (knots+2) splines functions for the approximation of the hazard or the survival functions. Rondeau, et al. (2012) suggested that the number of knots must be between 4 and 20. The default is 10.
a positive smoothing parameter value for the penalized likelihood estimation. The defaults is "10000".
a character string specifying prediction method applied on model. The possible options are "cox" for the classical Cox; "marg" for marginal and "cond" conditional prediction methods on shared models. The default is "cond".
a character string to state the distribution of random effect: "Gamma" for a gamma distribution, "LogN" for a log-normal distribution. Default is "Gamma".
types of hazard functions: "Splines" represents a semi-parametric hazard function using equidistant intervals and is estimated via penalized likelihood, "Splines-per" uses percentiles instead of equidistant intervals, "Piecewise-per" is a piecewise constant hazard function based on percentiles, "Piecewise-equi" is a piecewise constant hazard using equidistant intervals, and "Weibull" is a fully parametric hazard function based on the Weibull distribution. "Splines" is used as the default setting.
maximum number of iterations. The default is 300.
Beyene, K. M., and Chen, D. G. (2024). Time-dependent receiver operating characteristic curve estimator for correlated right-censored time-to-event data. Statistical Methods in Medical Research, 33(1), 162-181.
Beyene, K.M. and El Ghouch A. (2020). Smoothed time-dependent receiver operating characteristic curve for right censored survival data. Statistics in Medicine. 39: 3373-3396.
Rondeau, V., Marzroui, Y., & Gonzalez, J. R. (2012). frailtypack: an R package for the analysis of correlated survival data with frailty models using penalized likelihood estimation or parametrical estimation. Journal of Statistical Software, 47, 1-28.