[R] Creating factors from continuous variables

[Prof Brian Ripley]

?cut

This is in `An Introduction to R', the manual which ships with R and basic 
reading.

On Fri, 26 Aug 2005, David James wrote:

> What is the quickest way to create many categorical variables
> (factors) from continuous variables?
>
> This is the approach that I have used:
>
> # create sample data
> N <- 20
> x <- runif(N,0,1)
>
> # setup ranges to define categories
> x.a <- (x >= 0.0) & (x < 0.4)
> x.b <- (x >= 0.4) & (x < 0.5)
> x.c <- (x >= 0.5) & (x < 0.6)
> x.d <- (x >= 0.6) & (x < 1.0)
>
> # create factors
> i <- runif(N,1,1)
> x.new <- (i*1*x.a) + (i*2*x.b) + (i*3*x.c) + (i*4*x.d)
> x.factor <- factor(x.new)
>
> I'm looking for a better / simpler / more elegant / more robust (as
> the number of categories increases) way to do this.  I also don't
> like that my factor names can only be numbers in this example.  I
> would prefer a solution to take a form like the following (inspired
> by the "hist" function):
>
> # define breakpoints
> x.breaks = c(0, 0.4, 0.5, 0.6, 1.0)
> x.factornames = c( "0 - 0.4", "0.4 - 0.5", "0.5 - 0.6", "0.6 - 1.0" )
> x.factor = unknown.function( x, x.breaks, x.factornames )
>
> Thanks,
> David
>
> P.S. Here's what I have read to try to find the answer to my problem:
> * "Introductory Statistics with R"
> * "A Brief Guide to R for Beginners in Econometrics"
> * "Econometrics in R"

-- 
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