[R] about Cox-Box transformation

Prof Brian Ripley ripley at stats.ox.ac.uk
Thu Feb 9 13:59:15 CET 2006


On Thu, 9 Feb 2006, Petr Pikal wrote:

> Hi
>
> 1. 	upgrade
> 2.	read Introduction to R shipped with the distribution
> 3.	read documentation to MASS
> 4.	packages not libraries
> 5.	use library(MASS) to make functions from MASS package available
> 6. 	now you can use boxcox() but not on vectors AFAIK

You can by

boxcox(y ~ 1)

but car's box.cox does similar things.  (I did not mention that as adding 
packages to R 1.8.0 is unlikely to be easy.)

> probably not completely correct but this
>
> box.cox <- function(x, lambda,inv=F)
> {	if (!inv)
> 		{if(missing(lambda))
> 			log(x)
> 		else (x^lambda - 1)/lambda}
> 	else (lambda*x+1)^(1/lambda)
> }
>
> you can use to perform Box-Cox transformation on vectors.
>
> HTH
> Petr
>
>
> On 9 Feb 2006 at 11:51, Stefano Sofia wrote:
>
> From:           	Stefano Sofia <stefano.sofia at regione.marche.it>
> To:             	r-help at stat.math.ethz.ch
> Date sent:      	Thu, 09 Feb 2006 11:51:02 +0000
> Subject:        	[R] about Cox-Box transformation
>
>> Dear R-users,
>> I am using R version 1.8.0-1 under Suse 8.2.
>> I need to use the boxcox command because I want to apply a Cox-Box
>> transformation to a vector of rainfall values. Within the libraries,
>> the MASS library is present, but I don't know whether this means that
>> is automatically installed or not. The command doesn't work. What do I
>> have to do in order to make it working? Is just a problem of
>> installation or simply of recalling the right command?
>>
>> thank you for your help
>> Stefano

-- 
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 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595




More information about the R-help mailing list