Hi there, I have a question on fitting data by coxme. In particular I want to fit a random intercept and random slope cox model. Using the rats dataset as an example, I generated another covariate x2 and want to specify a random slope for x2. Here is my code: x2=matrix(rep(runif(50), 3), 50, 3) x2=as.vector(t(x2)) rats2=cbind(rats, x2) But when I used the coxme function as follows, it gave an error message. What is the right way to do it? Thanks a lot! coxme(Surv(time, status) ~ rx+x2, data=rats2, random=~ (1+x2)|litter ) Lei Liu Assistant Professor Division of Biostatistics and Epidemiology Dept. of Public Health Sciences School of Medicine University of Virginia 3181 Hospital West Complex Charlottesville, VA 22908-0717 1-434-982-3364 (o) 1-434-806-8086 (c) liulei at virginia.edu ll9f at virginia.edu
When I tried it, I seemed to get inconsistent results, the first of
which was irreproducible. The first error message said that 'rats2' was
not a data.frame. I don't know what I did to get that, but when I tried
it again, I got the following:
Error in max(kindex) : object "kindex" not found
> traceback()
3: max(kindex)
2: max(kindex)
1: coxme(Surv(time, status) ~ rx + x2, data = rats2, random = ~(1 +
x2) | litter)
This sounds to me like a logic error in 'coxme'; I am therefore
cc-ing the maintainer listed in help(package='kinship').
If it were my problem, I might list coxme and make a local copy of
it. It's easy to find the offending 'max(kindex)'. It's harder
to
figure out what kindex should be in this context. To attempt that, I
might request debug(coxme), then try again the 'coxme' call that
produced the error message and trace through it line by line until I
figured it out.
Hope this helps.
Spencer Graves
Lei Liu wrote:> Hi there,
>
> I have a question on fitting data by coxme. In particular I want to fit a
> random intercept and random slope cox model. Using the rats dataset as an
> example, I generated another covariate x2 and want to specify a random
slope
> for x2. Here is my code:
>
> x2=matrix(rep(runif(50), 3), 50, 3)
> x2=as.vector(t(x2))
>
> rats2=cbind(rats, x2)
>
> But when I used the coxme function as follows, it gave an error message.
> What is the right way to do it? Thanks a lot!
>
> coxme(Surv(time, status) ~ rx+x2, data=rats2, random=~ (1+x2)|litter )
>
>
> Lei Liu
> Assistant Professor
> Division of Biostatistics and Epidemiology
> Dept. of Public Health Sciences
> School of Medicine
> University of Virginia
>
> 3181 Hospital West Complex
> Charlottesville, VA 22908-0717
>
> 1-434-982-3364 (o)
> 1-434-806-8086 (c)
>
> liulei at virginia.edu
> ll9f at virginia.edu
>
> ______________________________________________
> 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
Many thanks. I am actually able to check for this with S-PLUS 7 for
Linux and obtained the same error message concerning kindex in coxme. It
is thus worthwhile to give a note to the S-PLUS author before looking
into the details. Best, JH
-----Original Message-----
From: Spencer Graves [mailto:spencer.graves at pdf.com]
Sent: 03 July 2006 17:56
To: Lei Liu
Cc: r-help at stat.math.ethz.ch; Jing hua Zhao
Subject: Re: [R] help with coxme
When I tried it, I seemed to get inconsistent results, the
first of
which was irreproducible. The first error message said that 'rats2' was
not a data.frame. I don't know what I did to get that, but when I tried
it again, I got the following:
Error in max(kindex) : object "kindex" not found
> traceback()
3: max(kindex)
2: max(kindex)
1: coxme(Surv(time, status) ~ rx + x2, data = rats2, random = ~(1 +
x2) | litter)
This sounds to me like a logic error in 'coxme'; I am
therefore
cc-ing the maintainer listed in help(package='kinship').
If it were my problem, I might list coxme and make a local
copy of
it. It's easy to find the offending 'max(kindex)'. It's harder
to
figure out what kindex should be in this context. To attempt that, I
might request debug(coxme), then try again the 'coxme' call that
produced the error message and trace through it line by line until I
figured it out.
Hope this helps.
Spencer Graves
Lei Liu wrote:> Hi there,
>
> I have a question on fitting data by coxme. In particular I want to
fit a > random intercept and random slope cox model. Using the rats dataset as
an > example, I generated another covariate x2 and want to specify a random
slope > for x2. Here is my code:
>
> x2=matrix(rep(runif(50), 3), 50, 3)
> x2=as.vector(t(x2))
>
> rats2=cbind(rats, x2)
>
> But when I used the coxme function as follows, it gave an error
message. > What is the right way to do it? Thanks a lot!
>
> coxme(Surv(time, status) ~ rx+x2, data=rats2, random=~ (1+x2)|litter
)>
>
> Lei Liu
> Assistant Professor
> Division of Biostatistics and Epidemiology
> Dept. of Public Health Sciences
> School of Medicine
> University of Virginia
>
> 3181 Hospital West Complex
> Charlottesville, VA 22908-0717
>
> 1-434-982-3364 (o)
> 1-434-806-8086 (c)
>
> liulei at virginia.edu
> ll9f at virginia.edu
>
> ______________________________________________
> 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