Dear r-helpers, This is my first time to run survival analysis. Currently, I have a data set which contains two variables, the variable of time to event (or time to censoring) and the variable of censor indicator. For the indicator variable, it was coded as 0 and 1. 0 represents right censor, 1 means event of interest. Now I try to use "survfit" in the package of "survival". I wrote the following code:> rptsurv <- survfit(surv(time,censor)~1,data=x)Before I run the code, I am concerned with my 0/1 coding to the censor indicator because I did not see any argument in the syntax of "survfit", which may tell the program that value 1 means event. I checked the documentations and R-help archive, but ended in vain. Would you please kindly tell me how "survfit" treats censor variables? In 0/1 coding, is it the default that 1 means event and 0 means right censor? What if the censor was coded as 2 or 3 instead of 0 or 1? I means how the "survfit" knows the difference. In SAS, if a "lifetest" procedure (similar to survfit) is performed, there is an argument specifying which value in the censor variable is treated as event. I know I could just compare the results from R and from SAS to see the difference. However, I really want to know exactly how "survfit" deals with this problem. Thank you very much in advance. sincerely, Jiang Lu University of Pittsburgh
The Surv object contains the information on the type of censoring. Look at ?Surv for an explanation of how censored events are represented. -Christos> -----Original Message----- > From: r-help-bounces at stat.math.ethz.ch > [mailto:r-help-bounces at stat.math.ethz.ch] On Behalf Of Lu, Jiang > Sent: Saturday, April 28, 2007 11:10 PM > To: r-help at stat.math.ethz.ch > Subject: [R] how to code the censor variable for "survfit" > > Dear r-helpers, > > This is my first time to run survival analysis. Currently, I > have a data set which contains two variables, the variable of > time to event (or time to censoring) and the variable of > censor indicator. For the indicator variable, it was coded as > 0 and 1. 0 represents right censor, 1 means event of > interest. Now I try to use "survfit" in the package of > "survival". I wrote the following code: > > rptsurv <- survfit(surv(time,censor)~1,data=x) > > Before I run the code, I am concerned with my 0/1 coding to > the censor indicator because I did not see any argument in > the syntax of "survfit", which may tell the program that > value 1 means event. I checked the documentations and R-help > archive, but ended in vain. > > Would you please kindly tell me how "survfit" treats censor variables? > In 0/1 coding, is it the default that 1 means event and 0 > means right censor? What if the censor was coded as 2 or 3 > instead of 0 or 1? I means how the "survfit" knows the > difference. In SAS, if a "lifetest" > procedure (similar to survfit) is performed, there is an > argument specifying which value in the censor variable is > treated as event. > > I know I could just compare the results from R and from SAS > to see the difference. However, I really want to know exactly > how "survfit" deals with this problem. Thank you very much in advance. > > sincerely, > > Jiang Lu > University of Pittsburgh > > ______________________________________________ > R-help at 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 > and provide commented, minimal, self-contained, reproducible code. > >
And be careful - R is case-sensitive. You have "surv(...)" instead of "Surv(...)" in your code, that will probably give an error. The coding is as you have it - 1=failure, 0=censored. Petr Christos Hatzis napsal(a):> The Surv object contains the information on the type of censoring. > Look at ?Surv > for an explanation of how censored events are represented. > > -Christos > > >> -----Original Message----- >> From: r-help-bounces at stat.math.ethz.ch >> [mailto:r-help-bounces at stat.math.ethz.ch] On Behalf Of Lu, Jiang >> Sent: Saturday, April 28, 2007 11:10 PM >> To: r-help at stat.math.ethz.ch >> Subject: [R] how to code the censor variable for "survfit" >> >> Dear r-helpers, >> >> This is my first time to run survival analysis. Currently, I >> have a data set which contains two variables, the variable of >> time to event (or time to censoring) and the variable of >> censor indicator. For the indicator variable, it was coded as >> 0 and 1. 0 represents right censor, 1 means event of >> interest. Now I try to use "survfit" in the package of >> "survival". I wrote the following code: >>> rptsurv <- survfit(surv(time,censor)~1,data=x) >> Before I run the code, I am concerned with my 0/1 coding to >> the censor indicator because I did not see any argument in >> the syntax of "survfit", which may tell the program that >> value 1 means event. I checked the documentations and R-help >> archive, but ended in vain. >> >> Would you please kindly tell me how "survfit" treats censor variables? >> In 0/1 coding, is it the default that 1 means event and 0 >> means right censor? What if the censor was coded as 2 or 3 >> instead of 0 or 1? I means how the "survfit" knows the >> difference. In SAS, if a "lifetest" >> procedure (similar to survfit) is performed, there is an >> argument specifying which value in the censor variable is >> treated as event. >> >> I know I could just compare the results from R and from SAS >> to see the difference. However, I really want to know exactly >> how "survfit" deals with this problem. Thank you very much in advance. >> >> sincerely, >> >> Jiang Lu >> University of Pittsburgh >> >> ______________________________________________ >> R-help at 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 >> and provide commented, minimal, self-contained, reproducible code. >> >> > > ______________________________________________ > R-help at 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 > and provide commented, minimal, self-contained, reproducible code. >-- Petr Klasterecky Dept. of Probability and Statistics Charles University in Prague Czech Republic