topic Re: How do I force the largest observe time to be always failure in SAS Programming

How do I force the largest observe time to be always failure

desireatem — Mon, 15 Jul 2013 15:20:55 GMT

Hello, in the following data generation; I want to change the censored value for the highest observed time to be always 0 '. That is after sorting by time from minumum time to maximum, I want censored=0 for the largest time

data test;

seed=-1;

alpha1 = 2.50;

beta1 = 9.50;

do i = 1 to 100;

lambdaT = 0.025; *baseline hazard;

lambdaC= .03; *light=0.15; *heavy=0.03;

er=0+sqrt(0.0001)*rannor(1);

t = rand("WEIBULL", 0.75, lambdaT); * time of event;

c = rand("WEIBULL", 1.25, lambdaC) ;* time of censoring;

time = min(t, c); * which came first?;

censored = (c lt t);

obs=(t lt c); * creating observation variable from censored when

y= alpha1 + beta1*t + er;

output;

end;

run;

Re: How do I force the largest observe time to be always failure

esjackso — Mon, 15 Jul 2013 16:29:20 GMT

Is this what you need?

Proc sort data=test out=test2; by descending time; run;

data test3;

set test2;

by descending time;

if first.time then censored = 0;

run;

Re: How do I force the largest observe time to be always failure

Astounding — Mon, 15 Jul 2013 16:35:46 GMT

I think the intent is simpler than that. The way I read it, the IF/THEN statement should be:

if _n_=1 then censored=0;

All of the other work is necessary ... sorting, then reading the data back into a DATA step. There is no easy way to make all of this happen in a single DATA step.

Re: How do I force the largest observe time to be always failure

PGStats — Mon, 15 Jul 2013 16:38:34 GMT

You can do it this way:

proc sql;

create table testMax as select max(t) as maxt from test;

update test

set censored = 0

where t = (select maxt from testMax);

drop table testMax;

quit;

Re: How do I force the largest observe time to be always failure

esjackso — Mon, 15 Jul 2013 16:40:18 GMT

Yeah in this case this two methods are the same.

Re: How do I force the largest observe time to be always failure

esjackso — Mon, 15 Jul 2013 16:45:08 GMT

I like this method (one correction I believe it not just t but min of t or c that makes up time) .. .you could probably select the obs of the max time into a macro variable and use that in the update where and avoid create the temp table all together.

Re: How do I force the largest observe time to be always failure

Astounding — Mon, 15 Jul 2013 16:55:03 GMT

PG,

I suspect you're fine with that ... Not knowing anything about a Weibull distribution, I wasn't sure if a tie would be possible. I didn't want to allow the possibility of uncensoring two observations. That's probably impossible, given the use of RAND, but I was too lazy to look it up when I had a working approach!

Re: How do I force the largest observe time to be always failure

PGStats — Mon, 15 Jul 2013 17:04:50 GMT

Well, OP's request wasn't that clear to me. The Weibull distribution is continuous, so that ties are very unlikely. I'm waiting for some feedback from OP.

Re: How do I force the largest observe time to be always failure

Reeza — Mon, 15 Jul 2013 17:29:57 GMT

You probably don't need to be generating a censoring time then do you? About half the lines in your original code could be deleted.