GitHub

SurvivalModels

Documentation for SurvivalModels.

In this documentation, we can cite stuff from the references.bib file like that : [1].

This documentation shall cover all the content of the package, which is not the case yet.

SurvivalModels.CoxModel — Type
CoxModel(init, times, status, des, method, maxit)
fit(CoxModel, @formula(Surv(T,Δ)~predictors), dataset, method, maxit)

Maximum Partial likelihood estimation in the semiparametric Cox Proportional Hazards model. W recomand using the formula interface.

  • times : times to event

  • status: vital status indicators (true or 1 = observed, false or 0 = censored)

  • des: design matrix for hazard-level effects

  • method: An algorithm from Optim.jl, e.g. NelderMead, Newton, LBFGS, ConjugateGradient or GradientDescent. For the moment our implementation only supports Optim.jl, but a switch to Optimizations.jl will be done in a second step.

maxit: maximum number of iterations in "method"

source
[1]
D. R. Cox. Regression models and life-tables. Journal of the Royal Statistical Society: Series B (Methodological) 34, 187–202 (1972).