[R] Design package: plot summary

Frank E Harrell Jr f.harrell at vanderbilt.edu
Wed Oct 3 00:12:44 CEST 2007


Cary Dehing-Oberije wrote:
> Hi everybody,
> 
> I am a new user of R, design package.
> I am trying to plot the estimated hazard ratio's of my cox regression
> model with the confidence intervals. But I keep getting the
> message:Error in `contrasts<-`(`*tmp*`, value = "contr.treatment") : 
>         contrasts can be applied only to factors with 2 or more levels
> 
> I have dichotomous, categorical as well as continuous variables in the
> model, but I can only plot the dichotomous variable.
> 
> I started with the following syntax:
> 
> dd <- datadist(x1, x2, x3)

Please follow the posting guide.  Please provide R commands that 
generate values of x1, x2, x3 that cause the problem you reported below.

> options(datadist='dd')
> 
> fit<-cph(S~ x1 +x2 +x3, x=T,y=T)
> 
> plothr<- summary(fit)
> plot(plothr, log=T)

or just plot(summary(fit), log=TRUE)

> 
> Because the levels defined by datadist didn't seem to work I tried to
> adjust the levels after fitting the model:
> 
> dd$limits["Adjust to","x1"] <- 1 
> dd$limits["Adjust to","x2"] <- 30

Don't do that.

Frank


> fit <- update(fit)
> plothr<- summary(fit)
> plot(plothr, log=T)
> 
> Finally I have tried: summary(fit, x1=1, x2=30, x3=3)
> 
> But this option didn't work either. Can anybody tell me what goes wrong
> and why?
> Thanks for your help,
> 
> Cary
> 
> Cary Dehing-Oberije
> Researcher CAT project
> p/a MAASTRO clinic
> Dr. Tanslaan 12
> 6229 ET Maastricht
> The Netherlands
> phone: 088 - 4455666 / 088 - 4455767
> email: cary.dehing at maastro.nl 
> 
> 
> 
> 
> 
> 
> 
> 	[[alternative HTML version deleted]]
> 
> ______________________________________________
> R-help at r-project.org 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.
> 


-- 
Frank E Harrell Jr   Professor and Chair           School of Medicine
                      Department of Biostatistics   Vanderbilt University



More information about the R-help mailing list