Indermaur Lukas wrote:
> Hi,
> I would like to evaluate the frequency of the variables within the best 
> selected model by AIC among a set of 12 competing models (I fit them with 
> GLM) with a bootstrap procedure  to get unbiased results. So I would ike to 
> do the ranking of the 12-model-set 10'000 times separately and calculate the 
> frequency of variables of the 10'000 best ranked models. I wrote a script 
> doing the model ranking (with some difficulty) for the pre-defined 
> 12-model-set, that works. My model results (one row per model) are stored in 
> a dataframe called "mydataframe". How can I implement my script simply into a 
> bootstrap and sum up the frequencies of variables of the 10'000 bootstraped 
> results?
>  
> I am not so good in R and would appreciate any hint.
>  
> Regards
> Lukas

I'm not entirely clear on your goals but the bootstrap selection 
frequency is mainly a replay of the significance of variables in the 
full model.  I don't see that it provides new information.  And 
selection frequency is marred by collinearity.

Frank

>  
>  
> °°° 
> Lukas Indermaur, PhD student 
> eawag / Swiss Federal Institute of Aquatic Science and Technology 
> ECO - Department of Aquatic Ecology
> Überlandstrasse 133
> CH-8600 Dübendorf
> Switzerland
>  
> Phone: +41 (0) 71 220 38 25
> Fax    : +41 (0) 44 823 53 15 
> Email: [EMAIL PROTECTED]
> www.lukasindermaur.ch
> 
> ______________________________________________
> [email protected] 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

______________________________________________
[email protected] 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.

Reply via email to