[R] Survival Regression with multiple records per subject

[Fabian Hefner]

Dear R users!

I reformulate the question with another example perhaps my question will be
more clearly now.
 
I have several subjects. One subject has multiple records. Only a starting
point exists the end point is vague.
Here is an example:
 
   itm      ID     exercise      time
1.401869    1        0             1
1.324390    1        0             2
1.324390    1        0             3
1.333338    1        0             4
1.346761    1        0             5
1.315441    1        1             6
1.337812    2        0             1
1.319915    2        0             2
1.351235    2        1             3
itm is the covariate;
ID is the subject Id;
exercise indicates if the subject is dead=1 or alive=0
 
How can I allocate the multiple records to one subject (for example record
1-6 are part of subject with ID 1 record 7-9 are part of subject with ID2)
and process a survival regression.
 
the "survRegData <- survreg(formula=Surv(time,exercise)~itm, data=Data,
dist="weibull")" command doesn't take into account that multiple records are
part of one subject.
 
Many thanks!
 
Fabian Hefner