NEWS

intsurv 0.3.0 (2025-09-29)

Simplified the interface of functions cox_cure() and cox_cure_net() through helper functions cox_cure.control(), cox_cure.mstep(), cox_cure_net.penalty().

Avoid unnecessary internal data standardization to improve computation efficiency.

Added AIC to outputs of fitted Cox cure models.
Added a new argument named cv_nfolds to cox_cure_net() and cox_cure_net.fit() for model selection by cross-validation.
Added new arguments named surv_offset and cure_offset for optional offset terms in survival model component and incidence model component, respectively, to cox_cure(), cox_cure.fit(), cox_cure_net(), and cox_cure_net.fit(). One or more offset terms can be also specified via corresponding formula.

Fixed issue caused by zeros in specified surv_l1_penalty_factor and cure_l1_penalty_factor for cox_cure_net() and cox_cure_net.fit(). Thank Sy Han Chiou for reporting this issue.

Moved C++ header files to inst/include/ so that they can be linked by other packages.
Added testing suite with help of the tinytest package.
Adjusted default maximum number of iterations in M-steps for possibly faster convergence in function cox_cure and cox_cure_net.

Added function cox_cure for fitting regular Cox cure rate model and an extended Cox cure rate model for right-censored data with uncertain event status.
Added function cox_cure_net for fitting regularized Cox cure rate model for right-censored data with possible uncertain event status.
Added function cIndex for computing concordance index with possible weights.
Added function simData4cure for simulating survival data with possible uncertain event status.

Alpha version of the package for paper submission and ease of reproducing main simulation results.