Determines the number of boosting steps for a survival model fitted by CoxBoost via cross-validation, conforming to the calling convention required by argument complexity in peperr call.
complexity.mincv.CoxBoost(response, x, full.data, ...)a survival object (Surv(time, status)).
n*p matrix of covariates.
data frame containing response and covariates of the full data set.
additional arguments passed to cv.CoxBoost call.
Scalar value giving the optimal number of boosting steps.
Function is basically a wrapper around cv.CoxBoost of package CoxBoost. A K-fold cross-validation (default K=10) is performed to search the optimal number of boosting steps, per default in the interval (0, maxstepno=100). The number of boosting steps with minimum mean partial log-likelihood is returned. Calling peperr, the default arguments of cv.CoxBoost can be changed by passing a named list containing these as argument args.complexity.
peperr, cv.CoxBoost