surv_sims: Function to simulate survival data.
In rOpenHealth/rEHR: Manipulating and Analysing Electronic Health Record Data

Description Usage Arguments Details Value

Model: proportional hazards, h(t; cov_mat, beta) = exp(cov_mat indicators for Type I censoring (common censoring time 'tc').

1
2
3

surv_sims(cov_mat, beta, cens_type = c("typeI", "noninformative"),
  baseline_hazard, cens_hazard = 0.04, cens_prob = 0, scale = 1,
  weibull_shape = 1)

`cov_mat`	n x p matrix of cov_matiates
`beta`	p-vector of regression coefficients
`cens_type`	typeI censoring or non-informative based on exponential distribution
`baseline_hazard`	for modelling death dates
`cens_hazard`	log(hazard) for non-informative censoring
`cens_prob`	expected censoring fraction (0 <= cens_prob < 1). Used for typeI censoring
`scale`	value to scale up the time variable by
`weibull_shape`	shape parameter for the weibull distribution. 1 is the same as an exponential

Weibull_shape is the k (shape) parameter from a weibull distribution

A value of k < 1 indicates that the mortality rate decreases over time. This happens if there is significant infant mortality
A value of k = 1 indicates that the mortality rate is constant over time. This might suggest random external events are causing mortality. This is the same as an exponential distribution
A value of k > 1 indicates that the mortality rate increases with time. This happens if there is an aging process.

Censored exponential survival times and censoring

rOpenHealth/rEHR documentation built on May 26, 2019, 8:51 p.m.