02-11-2013 12:12 PM
The following code gives me the estimated survival function for just the events, How can I get the baseline survival for both events and censored observation.
I need the estimated baseline survival'
proc phreg data=ph;
model zz*status(0)=x1/RL TIES=EFRON;
baseline out=survs survival=s;
02-11-2013 12:46 PM
I think both censored and event have the same baseline estimates based on the cox model.
The baseline function just happens to give estimates at event times, rather than censored times.
You can specify the timelist option if you want the estimate for specific time intervals instead of event times.
Check the documentation on how to specify the time list for baseline.
02-11-2013 01:03 PM
I'm fairly certain that the baseline should be the same for censored and uncensored.
Can you provide sample stata code that you want converted or preferably a link to their documentation that show how it is calculated.