[R] e1071 tune.control() random parameter

Jessica Streicher jstreicher at micromata.de
Tue Apr 3 14:49:36 CEST 2012


The function is a parameter to the tune function of library e1071. The 
other parameters of the tune.control() function specify more or less the 
sampling method (cross-validation, fixed or bootstrap+attributes for those)

tune itself iterates over valuevectors of the parameters (e.g. C and 
gamma) and calculates errors in the classification for each combination 
(using the provided sampling method in tune.control() i would guess)

also: e1071 = libsvm
and i have enough problems with R, i definetly wont start with python now ;)

The question is rather of "academic" interest, i can do what i want to 
do already. But it COULD be something useful, after all it is the first 
argument of the function^^

On 03.04.12 14:13, Alekseiy Beloshitskiy wrote:
> Hello, Jessica,
>
> Can you please elaborate what you're trying to find with that function.
> If e.g. you want to find best parameters C and gamma for RBF model (SVM), you can use grid.py, check here:
> http://www.csie.ntu.edu.tw/~cjlin/libsvm/
>
> Function tune.control() in package e1071 is an R interface to this algorithm.
>
> Best,
> -Alex
> ________________________________________
> From: r-help-bounces at r-project.org [r-help-bounces at r-project.org] on behalf of Jessica Streicher [j.streicher at micromata.de]
> Sent: 03 April 2012 11:05
> To: r-help at r-project.org
> Subject: [R] e1071 tune.control() random parameter
>
> I'm not sure what the parameter specifies:
>
> random
> if an integer value is specified, random parameter vectors are drawn from the parameter space.
> What are the parameter vectors and what is the parameter space? What means drawn?
>
> greetings
> Jessi
>          [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org 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.



More information about the R-help mailing list