[R] weibull in R

[Prof Brian Ripley]

On Thu, 17 Jan 2002, Ronaldo Reis Jr. wrote:

> Hi all
> I try to make a weibull survival analysis on R.
> I know make this on GLIM, and now I try to make the GLIM exercice GLEX8 on R
> to learning and compare the test.
> The variables are:
> time censor group bodymass
>
> In GLIM I make:
> $calc %s=1 $ to fit weibull rather than exponential
> $input %pcl weibull $
> $macro model group*bodymass $endmac$
> $use weibull t w %s $
>
> Then, GLIM estimate an alpha for the model group*bodymass, the value on GLIM
> is 0.736.
>
> I try to make this on R.
>
> > survreg(Surv(time,censor)~group*bodymass,dist="weibull",scale=0)

You don't need to give default values in R.

> The result is an alpha (Scale on R) different.
>
> Scale= 1.36

1/1.36 is 0.7352941

It's parameterized differently.  Whatever `alpha' means, it is not what
the survival package is using as the Scale parameter in survreg.

> Where is the error????

Nowhere.  There are lots of parameter sets for the Weibull, so you have to
check each package to see what it uses.  And as you are using an addon-on
package and not R, I do mean package survival, and not R (where
dweibull is different).

> Exist any reference to learning survival in R?

This is covered in the appropriate chapter of Venables & Ripley.
(On page 375.)

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._