On Dec 19, 2013, at 5:32 AM, L... L... wrote:
> Dear all, is there an R function to simulate random observations from a
cure fraction model (random observations with long-term survivos).
> Some references how can I do this will be welcome.
>
Shouldn't be too hard to set up a mixture of weibull distributed event times
with one component having a parameter less than "exponential" (to
represent the cured fraction with survival typical of an average human
population) and another with a hazard like an "exponential" or worse
to represent the non-cured fraction. Watch out about how the shape parameters
are defined. The survival package uses different parameterization than the
rweibull definitions.
library(survival)
?survreg
?rweibull
--
David Winsemius
Alameda, CA, USA