[R] AIC bug?

[Prof Brian D Ripley]

On Tue, 6 Mar 2001, Jean Lejeune wrote:

> Dear all,
>
> I am a little problem. In the help, AIC = - 2log L + k*edf. When the model
> is linear, the help said " -2log L is the deviance ". I have a model

Which help?  There is no saturated model and hence strictly no deviance for
a Gaussian linear model.  deviance(toto.lm) is the residual sum of squares,
not -2 log maximized L.

?glm says

deviance: up to a constant, minus twice the maximized log-likelihood.
          Where sensible, the constant is chosen so that a saturated
          model has deviance zero.

Is that what you are thinking of?

BTW, I think k=2 for Akaike's AIC: there are others....

> toto.lm  with one output and three input where
> deviance(toto.lm) = 8.027 and edf =4. But AIC = -31.55. I don't understand
> why?
> Many thanks.
>     Jean LEJEUNE Université de CAEN (France)
> -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
> 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
> _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
>

-- 
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._