Hi list,
I?m testing out uniCox R package (version 1.0, on R2.12.0, WinXP).
When I ran uniCox on my data, there are always some NA?s in the beta matrix,
which in turn causes problems in uniCoxCV call. I don?t see anything wrong
with the corresponding data (e.g. no NAs) and if I fit a univariate Cox model,
the features that give NA beta estimates are actually pretty significant.
Could you please let me know what happened and how to avoid this?
I?ve attached the outputs of the function calls below.
Thank you very much!
...Tao
> a <- uniCox(x=t(dat.ave.train.base),
y=sampleinfo.ave.train.base$tm2dthr,
>status=sampleinfo.ave.train.base$censrdth)
lambda value 1 out of 20
lambda value 2 out of 20
lambda value 3 out of 20
lambda value 4 out of 20
lambda value 5 out of 20
lambda value 6 out of 20
lambda value 7 out of 20
lambda value 8 out of 20
lambda value 9 out of 20
lambda value 10 out of 20
lambda value 11 out of 20
lambda value 12 out of 20
lambda value 13 out of 20
lambda value 14 out of 20
lambda value 15 out of 20
lambda value 16 out of 20
lambda value 17 out of 20
lambda value 18 out of 20
lambda value 19 out of 20
lambda value 20 out of 20
5 betas missing
> aa <- uniCoxCV(a, x=t(dat.ave.train.base),
>y=sampleinfo.ave.train.base$tm2dthr,
status=sampleinfo.ave.train.base$censrdth)
FOLD= 1
lambda value 1 out of 20
lambda value 2 out of 20
lambda value 3 out of 20
lambda value 4 out of 20
lambda value 5 out of 20
lambda value 6 out of 20
lambda value 7 out of 20
lambda value 8 out of 20
lambda value 9 out of 20
lambda value 10 out of 20
lambda value 11 out of 20
lambda value 12 out of 20
lambda value 13 out of 20
lambda value 14 out of 20
lambda value 15 out of 20
lambda value 16 out of 20
lambda value 17 out of 20
lambda value 18 out of 20
lambda value 19 out of 20
lambda value 20 out of 20
3 betas missing
1
Error in coxph(Surv(y[ii], status[ii]) ~ eta.new) :
No (non-missing) observations
> a[[2]][(rowSums(is.na(a[[2]])))>0,]
[,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12] [,13]
[,14] [,15] [,16] [,17] [,18] [,19] [,20]
[1,] 92.6641 NaN NaN NaN 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0
[2,] NaN 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0
[3,] 567.3650 NaN 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0