[R] cph() in Design package

[array chip]

sorry this problem only occurs in S-Plus, not in R.

--- array chip <arrayprofile at yahoo.com> wrote:

> Hi there, I encountered a weird problem using cph()
> with Design package:
> 
> I have 2 datasets, say "dat1" and "dat2", both data
> frames with 3 columns "time","status" and "scores",
> all numeric
> 
> If I run the following:
> dd<-datadist(dat1)
> options(datadist='dd')
> dd
> 
>                  time status   scores 
>      Low:effect  37.0      0   -6.018
>       Adjust to  88.0      0   11.252
>     High:effect 109.0      1   33.267
>  Low:prediction  11.0      0  -34.419
> High:prediction 147.5      1   88.218
>             Low   5.0      0  -72.072
>            High 171.0      1  154.678
> 
> fit <- cph(Surv(time,status) ~ scores,
> data=dat1,surv=T, x=T, y=T)
> fit
> 
> Cox Proportional Hazards Model
> 
> cph(formula = Surv(time, status) ~ scores, data =
> xx,
> x = T, y = T, surv = T)
> 
>  Obs Events Model L.R. d.f.      P Score Score P   
> R2
> 
>  171     64      10.65    1 0.0011 12.01  0.0005
> 0.062
> 
> so everything is ok.
> 
> However, if I run the following with the second
> dataset:
> 
> dd<-datadist(dat2)
> options(datadist='dd')
> dd
> 
>                   time status   scores 
>      Low:effect  68.10      0  -16.036
>       Adjust to  91.75      0    9.068
>     High:effect 119.03      1   42.447
>  Low:prediction  21.75      0  -48.103
> High:prediction 120.00      1   96.661
>             Low   6.90      0  -75.392
>            High 120.00      1  141.524
> 
> fit <- cph(Surv(time,status) ~ scores,
> data=dat2,surv=T, x=T, y=T)
> fit
> 
> Problem: Object "fit" not found, while calling
> subroutine S_agsurv2 
> Use traceback() to see the call stack
> 
> traceback()
> 4: eval(action, sys.parent())
> 3: doErrorAction("Problem: Object \"fit\" not found,
> while calling subroutine S_agsurv2", 1000)
> 2: eval(expression(fit))
> 1: 
> Message: Problem: Object "fit" not found, while
> calling subroutine S_agsurv2
> 
> so it didn't work with the second dataset. The 2
> dataset is very similar in terms of column data
> type. 
> 
> any suggestions?
> 
> Many thanks
> 
> 
> 
> 
> 
> 
> __________________________________________________
> Do You Yahoo!?

> protection around 
> http://mail.yahoo.com 
>