Jens - After reading Help("Surv"), yes, I think you have interpreted the two required arguments to Surv() correctly. I don't know of a ready-made function to do the transformation you illustrate.
If I had to program it myself, I would use diff(), rep() and order(). - tom blackwell - u michigan medical school - ann arbor - On Sun, 5 Oct 2003, Jens Schmidt wrote: > I have survival data from a population: > > time alive treatment > 0 20 A > 3 12 A > 5 10 A > 8 7 A > 10 7 A > 15 5 A > 20 5 A > > (time=0 is start of the experiment and initial number of individuals is > alive=20; time=20 is end of experiment) > > If I don't misunderstood the survival documentation, always a survival > object is required, i.e. created with function Surv(time, event) > > So I have to transform my data (assuming no censoring, event=0 is alive > and event=1 is dead): > > time event treatment > 20 0 A > 20 0 A > 20 0 A > 20 0 A > 20 0 A > 15 1 A > 15 1 A > 8 1 A > 8 1 A > 8 1 A > 5 1 A > 5 1 A > 3 1 A > 3 1 A > 3 1 A > 3 1 A > 3 1 A > 3 1 A > 3 1 A > 3 1 A > > Is there a function or any other [R] code doing this? > > or > > Is it possible to perform survival analysis direct from > population data set? > > Thanks for any help > Jens > -- > Jens Schmidt TU Dresden -Inst. f. Hydrobiologie- > Tel. +49-(0)351-463-36284 Mommsenstrasse 13 > Fax +49-(0)351-463-37108 01062 Dresden > R.F.A. ______________________________________________ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help