On Jun 8, 2013, at 4:58 AM, matteo pettinari wrote:
> Dear
> I would like to find a nice way to plot the predicted survival obtained
> from a cph fit and few covariates as Age etc...
>
> I am working with "predict" from rms package but i found a
problem after
> the lines
>
> predict(cox5, type="lp", se.fit=FALSE, conf.int=TRUE,
conf.type=c('mean'))
Many functions in rms woek better with Predict()
>
> Error in Getlimi(name[i], Limval, need.all = TRUE) :
> no limits defined by datadist for variable Age
You need to read:
?datadist
And then create a datyadist and don't do what I do all the time and for get
to set the options()
The code you offer does not have the cph call which would have been needed to
construct a proper datadist argument. Perhaps something like:
ddname <- datadist(yourdataframe)
options(datadist="ddname")
> --
>
>
> can someone give me any advice about this problem and to create nice plot
> with x axis a continuous predictor and y axis the predicted survival?
>
> Thank you
> Dr. Matteo Pettinari
> matteo.pettinari at gmail.com
> matteo.pettinari at uzleuven.be
>
> Tervuursestraat 43 | 3000 | Leuven | Belgium
>
> [[alternative HTML version deleted]]
Please learn to post in plain text.
David Winsemius
Alameda, CA, USA