[R] Quasi AIC

Uwe Ligges ligges at statistik.uni-dortmund.de
Fri Jul 4 23:16:18 CEST 2003


Henric Nilsson wrote:
> 
> Dear all,
> 
> Using the quasibinomial and quasipoisson families results in no AIC being
> calculated. However, a quasi AIC has actually been defined by Lebreton et al
> (1992). In the (in my opinon, at least) very interesting book by Burnham and
> Anderson (1998,2002) this QAIC (and also QAICc) is covered. Maybe this is something
> that could be implemented in R.
> 
> Take a look at page 23 in this pdf:
> 
> http://www.springer-ny.com/supplements/0387953647/Preface_Contents_and_Chapter_8.pdf
> 
> Lebreton, J-D et al. (1992) Modeling Survival and Testing Biological Hypotheses
> Using Marked Anomals: A Unified Approach with Case Studies. Ecological Monographs,
> 62(1), 67-118.
> 
> Burnham, K.P. & Anderson, D.R. (2002) Model Selection and Multi-Model Inference - A
> Practical Information-Theroretic Approach (2nd Ed). Springer Verlag.


Well, I tried to send a private message, but the e-mail address you have
specified seems to be unknown for your mail server ... Please fix it
before posting messages.

If you think QAIC is useful and there's no implemented solution yet (I
haven't found any in a *quick* search through CRAN packages), patches
and suggestions including code and documentation are welcome as usual (I
don't know that much about QAIC, and cannot judge its worth, though).

Uwe Ligges




More information about the R-help mailing list