- Added AIC to outputs of fitted Cox cure models.
- Added a new argument named
cv_nfolds
tocox_cure_net()
andcox_cure_net.fit()
for model selection by cross-validation. - Added new arguments named
surv_offset
andcure_offset
for optional offset terms in survival model component and incidence model component, respectively, tocox_cure()
,cox_cure.fit()
,cox_cure_net()
, andcox_cure_net.fit()
. One or moreoffset
terms can be also specified via corresponding formula.
- Fixed issue caused by zeros in specified
surv_l1_penalty_factor
andcure_l1_penalty_factor
forcox_cure_net()
andcox_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
andcox_cure_net
.
- Fixed weight indices in function
cIndex
.
- 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.
- Added methods for parameters' initialization.
- Alpha version of the package for paper submission and ease of reproducing main simulation results.