coxSimulate: Function for simulating a Cox distribution series
In kbolab/distrcox: Package for simulating Cox proportional hazards distributions

Description Usage Arguments Details Value

Function for simulating a Cox distribution series

1	coxSimulate(n = 1000, input.model, features, BH)

`n`	Number of cases to be simulated
`input.model`	An object of `coxph` class, even anonymized (see `coxAnonymize`)
`features`	An object of class `features` containing description of single covariates in the dataset
`BH`	Cumulative baseline hazard as calculated by `basehaz` function applied to a `coxph` object

Cox distribution series are based on the assumption of proportional hazards of covariate over a given underneath baseline hazards function. The hypotesis of proportionality is essential for achieving good performance by models generated by this kind of fitting. The coxSimulate function works on an input coxph model. The model can be also anonymized by coxAnonymize function, in order to use the simulation function in a distributed learning context, without moving data from their site of origin and without carrying details of single patients within the coxph object that might even hypothetically be used to reproduce the original dataset coming grom its own (far) source. Baseline hazard has to be given in order to produce a cox simulation because it cannot be anymore calculated by using the coxph

A data.frame containing simulated observation

kbolab/distrcox documentation built on May 20, 2019, 8:10 a.m.