R help - Feb 2003 - Splitting survivor episodes

To R-list --

Does anyone know of an R function that will create split-episode data from
single spell event/duration data according to user-defined time intervals?

Example: original data
 t     d  x
------------
 6     0  x1
 5     1  x2

Split using intervals [0,3) [3,infty) (or cutpoint at 3)

 start end    event
 _t0    _t     _d   _x  episode
   0     3      0   x1   1
   3     6      0   x1   2
   0     3      0   x2   1
   3     5      1   x2   2

Thanks,
Dan
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-Daniel A. Powers, Ph.D.
Associate Professor of Sociology
University of Texas at Austin
370 Burdine
Austin, TX  78712
phone: 512-232-6335
fax:   512-471-1748
dpowers at mail.la.utexas.edu

On Fri, 28 Feb 2003, Daniel A. Powers wrote:
> To R-list --
>
> Does anyone know of an R function that will create split-episode data from
> single spell event/duration data according to user-defined time intervals?
>
> Example: original data
>  t     d  x
> ------------
>  6     0  x1
>  5     1  x2
>
> Split using intervals [0,3) [3,infty) (or cutpoint at 3)
>
>  start end    event
>  _t0    _t     _d   _x  episode
>    0     3      0   x1   1
>    3     6      0   x1   2
>    0     3      0   x2   1
>    3     5      1   x2   2
>
No.

I do something like

   n<-nrow(data)
   data<-rbind(data,data)
   data$episode<-rep(c(1,2),each=n)
   data$start<-rep(c(0,CUT),each=n)
   data$end <- c(rep(CUT,n), data$t)

   ##drop episodes after death
   data<-subset(data, start<t )

   ## define event
   data$event <- with(data, ifelse(t<end, event, 0))

   ## tidy up end times
   data$end <- pmin(data$t, data$end)


	-thomas