R help - Oct 2004 - Making a 'joint distribution'?

Suppose I make two discrete variables --
> D <- data.frame(f1=sample(1:5,100,replace=T),
f2=sample(1:5,100,replace=T))
I know I can do:
> table(D$f1, D$f2)
    0 1 2 3 4
  0 5 5 5 5 4
  1 4 2 6 7 3
  2 5 3 5 3 6
  3 3 1 3 1 2
  4 6 4 3 3 6
> table(D$f1)
 0  1  2  3  4 
24 22 22 10 22 
> table(D$f2)
 0  1  2  3  4 
23 15 22 19 21 

which is all great. But how do I produce the typical presentation of
the "joint distribution" where we put the marginal distributions in
the margins? E.g. I'd like to get some object "joint" where one
would
get :
> joint
     0  1  2  3  4  f1
  0  5  5  5  5  4  24
  1  4  2  6  7  3  22
  2  5  3  5  3  6  22
  3  3  1  3  1  2  10
  4  6  4  3  3  6  22
 f2 23 15 22 19 21 100

So that one could then say "joint/nrow(D)" and get nice probabilities
out of it. It would great to be able to say "xtable(joint/nrow(D))"
:-)

I'm sure R has a lovely way to do this, but I'm not able to figure it
out.

-- 
Ajay Shah                                                   Consultant
ajayshah at mayin.org                      Department of Economic Affairs
http://www.mayin.org/ajayshah           Ministry of Finance, New Delhi

Dear Ajay,

There may be a function that does this already, but if not, it not hard to
do, at least for two-way tables:

Table <- function(x, y) {
      tab <- table(x, y)
      tab <- cbind(tab, rowSums(tab))
      tab <- rbind(tab, colSums(tab))
      rownames(tab)[nrow(tab)] <- deparse(substitute(y))
      colnames(tab)[ncol(tab)] <- deparse(substitute(x))
      tab
      }

I hope this helps,
 John 
> -----Original Message-----
> From: r-help-bounces at stat.math.ethz.ch 
> [mailto:r-help-bounces at stat.math.ethz.ch] On Behalf Of Ajay Shah
> Sent: Sunday, October 03, 2004 4:10 AM
> To: r-help
> Subject: [R] Making a 'joint distribution'?
> 
> Suppose I make two discrete variables --
> > D <- data.frame(f1=sample(1:5,100,replace=T), 
> > f2=sample(1:5,100,replace=T))
> 
> I know I can do:
> 
> > table(D$f1, D$f2)
>     0 1 2 3 4
>   0 5 5 5 5 4
>   1 4 2 6 7 3
>   2 5 3 5 3 6
>   3 3 1 3 1 2
>   4 6 4 3 3 6
> > table(D$f1)
>  0  1  2  3  4
> 24 22 22 10 22 
> > table(D$f2)
>  0  1  2  3  4
> 23 15 22 19 21 
> 
> which is all great. But how do I produce the typical 
> presentation of the "joint distribution" where we put the 
> marginal distributions in the margins? E.g. I'd like to get 
> some object "joint" where one would get :
> 
> > joint
>      0  1  2  3  4  f1
>   0  5  5  5  5  4  24
>   1  4  2  6  7  3  22
>   2  5  3  5  3  6  22
>   3  3  1  3  1  2  10
>   4  6  4  3  3  6  22
>  f2 23 15 22 19 21 100
> 
> So that one could then say "joint/nrow(D)" and get nice 
> probabilities out of it. It would great to be able to say 
> "xtable(joint/nrow(D))" :-)
> 
> I'm sure R has a lovely way to do this, but I'm not able to 
> figure it out.
> 
> -- 
> Ajay Shah                                                   Consultant
> ajayshah at mayin.org                      Department of Economic Affairs
> http://www.mayin.org/ajayshah           Ministry of Finance, New Delhi
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide! 
> http://www.R-project.org/posting-guide.html

On Sun, 3 Oct 2004, Ajay Shah wrote:
> Suppose I make two discrete variables --
>> D <- data.frame(f1=sample(1:5,100,replace=T),
f2=sample(1:5,100,replace=T))
>
<sniP>
> which is all great. But how do I produce the typical presentation of
> the "joint distribution" where we put the marginal distributions
in
> the margins? E.g. I'd like to get some object "joint" where
one would
> get :
?addmargins.

 	-thomas

Thanks to everyone who helped me solve this question. My cleanest
solution is:

joint.and.marginals <- function(x,y) {
  t <- addmargins(table(x, y))
  rownames(t)[nrow(t)] <- deparse(substitute(y))
  colnames(t)[ncol(t)] <- deparse(substitute(x))
  return(t)
}

There are many other valid solutions, but this one struck me as being
the simplest.

As a demo of it's use:
> D <- data.frame(f1=sample(1:5,10000,replace=T),
f2=sample(1:5,10000,replace=T)
> system.time(print(joint.and.marginals(D$f1, D$f2)))
      y
x         1    2    3    4    5  D$f1
  1     420  427  385  376  423  2031
  2     425  432  429  375  347  2008
  3     405  419  434  401  352  2011
  4     374  374  370  417  403  1938
  5     403  381  409  388  431  2012
  D$f2 2027 2033 2027 1957 1956 10000
[1] 0.05 0.00 0.07 0.00 0.00

Hmm, how would one get rid of the 'x' and 'y' that are occuring
in the
table? :-)

	-ans.