[R] simple question with table()

[Peter Dalgaard]

Berwin A Turlach wrote:
> G'day Simone,
> 
> On Sun, 30 Nov 2008 11:05:13 +0100
> Simone Gabbriellini <simone.gabbriellini at gmail.com> wrote:
> 
>> my problem should be easy to fix, but I couldn't find a solution by  
>> myself...
>>
>> In my survey, there is a question with 14 possible answers. None of  
>> the respondents choose the 13th answer, so when I table() the
>> results, R says:
> [...]
>> 13 is missing... anyone knows how to tell table() that there are 14  
>> modalities in the answers?
> 
> The easiest way is probably to turn your data into a factor with the
> appropriate set of levels:
> 
> R> dat <- sample(c(1:12,14), 100, replace=TRUE)
> R> table(factor(dat, levels=min(dat):max(dat)))
> 
>  1  2  3  4  5  6  7  8  9 10 11 12 13 14 
>  9  5  4  7  7 11  8 10  9  8  6 11  0  5 
> 
> that was once the solution of one of my colleagues and I find it
> somewhat nicer than the one I came up with:
> 
> R> rng <- min(dat):max(dat)
> R> res <- colSums(outer(dat, min(dat):max(dat), "=="))
> R> names(res) <- rng
> R> res
>  1  2  3  4  5  6  7  8  9 10 11 12 13 14 
>  9  5  4  7  7 11  8 10  9  8  6 11  0  5 


Notice, though, that you don't always want to have the limits being 
data-dependent either. E.g. if you have multiple questions of the 
variety "on a scale of 1:5 how do you feel...", you presumably want your 
barcharts on the same scale even if some questions are all 1's or all 5's.

So the straightforward table(factor(x,levels=1:14)) might be preferable.


-- 
    O__  ---- Peter Dalgaard             Øster Farimagsgade 5, Entr.B
   c/ /'_ --- Dept. of Biostatistics     PO Box 2099, 1014 Cph. K
  (*) \(*) -- University of Copenhagen   Denmark      Ph:  (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)              FAX: (+45) 35327907