[R] Reporting with down and across variables

[hadley wickham]

On Thu, Sep 18, 2008 at 10:03 AM, Matthew Pettis
<matthew.pettis at gmail.com> wrote:
> [Reposting with changed example]
> Hi,
>
> I want to take the dataframe df generated below and reshape the data with
> column names being w, x, y, and the different levels of z.  The values under
> the different levels of z are the corresponding values of r.  I've tried
> reshape and cast, and I can't seem to find the right combination.  Any help
> is appreciated,
>
> Matt
> ---------------------------------
>   # Generate data
> w <- c("a","b")
> x <- c("c","d")
> y <- c("e","f")
> z <- c("g","h")
> df <- expand.grid(w=w,x=x,y=y,z=z)
> df$r <- rnorm(16,mean=0,sd=1)

Well if you change that last line to :

df$value <- rnorm(16,mean=0,sd=1)

Then I think the following does what you want:

cast(df, w + x + y ~ z)

Hadley

-- 
http://had.co.nz/