Check the literature - an excellent article is
"rhDNase as an example of recurrent Event Analysis, Therneau, Hamilton,
Statistics in Medicine, Vol 16, 2029-2047"
Implicitly the intervals are "( ... ]".
-----Original Message-----
From: r-help-bounces@stat.math.ethz.ch
[mailto:r-help-bounces@stat.math.ethz.ch] On Behalf Of singyee ling
Sent: Wednesday, May 17, 2006 8:01 AM
To: r-help@stat.math.ethz.ch
Subject: [R] question about survSplit
Dear R-users,
I use the survsplit function in the survival package to change my data into
counting-process format
and the transformed format is as follow:
(a)
start stop event DP age ....
0 5 0 1 20
5 10 0 1 20
10 25 1 1 20
looking at the above three entries that belong to the same person, if an
event happen at time 5, won't the person actually enter the risk set twice
since there is another entry that start at time 5 and Cox proportional
hazard model "won't know" that it actually belong to the same
person.
Shouldn't it be like this?
(b)
start stop event DP age ....
0 5 0 1 20
6 10 0 1 20
11 25 1 1 20
or the R-function coxph has already take this into account when calculating
its risk set although I believe (a) is actually the correct one.
Your advice is greatly appreciated!
regards,
sing yee
[[alternative HTML version deleted]]
______________________________________________
R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide!
http://www.R-project.org/posting-guide.html
[[alternative HTML version deleted]]