>I am interested in fitting a discrete hazard model with a long-term
>survival mixture, i.e. there is a portion of the units who will never
>experience the event. I wonder if any one has some SAS code/ examples to
>get me started.
There are a pile of examples for this. Go to the SAS Online Docs
and look up PROC LIFEREG in the SAS/STAT section. This is
all PROC LIFEREG does. It handles a wide variety of survival
distribution, and provides several ways of defining your censoring.
>Specifically, my data is already in the unit-time period pair format.
>id time event x1 x2
>1 1 0 1 2
>1 2 0 1 2
>1 3 1 1 2
>2 1 1 0 2
>3 1 0 2 2
>3 2 0 2 3
>3 3 0 2 1
>3 4 0 2 3
>For the discrete hazard part, I want to include x2 as a time-vary
>covariate, and for the mixture (long term survival part), I want to use x1
>as part of the logit function. Thanks a lot!
This should be just fine for PROC LIFEREG. But it's really not a good idea
to insist that the form of your database dictate your method of analysis.
It's way easier to fix the database structure in SAS.
I suggest that you read the PROC LIFEREG docs, starting with the
"Overview" and "getting Started" sections, followed by the section
on the MODEL statement and the many examples. That should get
And don't forget to think about diagnostic plots at the end!
David L. Cassell
3115 NW Norwood Pl.
Corvallis OR 97330
FREE pop-up blocking with the new MSN Toolbar � get it now!