[R] the dimname of a table

[Gabor Grothendieck]

On 6/23/05, Marc Schwartz <MSchwartz at mednetstudy.com> wrote:
> On Thu, 2005-06-23 at 23:12 +0800, ronggui wrote:
> > i have a data frame(dat) which has many variables.and i use the
> > following script to get the crosstable.
> >
> > >danx2<-c("x1.1","x1.2","x1.3","x1.4","x1.5","x2","x4","x5","x6","x7","x8.1","x8.2","x8.3","x8.4","x11",
> > "x13","x17","x19","x20","x21")
> > >indep<-c("x23","x24","x25","x26","x27","x28.1","x28.2","x29")
> > >for (k in indep){
> >   for (i in danx2){
> >      a<-chisq.test(dat[,i],dat[,k])$p.v<=0.05
> >      if (a)
> > {CrossTable(dat[,i],dat[,k],chisq=T,format="SPSS");cat(rep("=",50),"\n","\n")}
> >  }
> >
> >   it has a little pitfall:the dimnames of table is dat[,i] and
> > dat[,k],but i want it to be like x2,x23...
> > is there any good way to do this?
> >   and in the command CrossTable(dat[,i],dat[,k],chisq=T,format="SPSS")
> > in the loop,is there any other way to get the variable other than
> > dat[,i] and dat[,k]?
> >   thank you !
> 
> Hi,
> 
> I am in between meetings here. Sorry for the delay in my reply to your
> query.
> 
> The best solution is for me to add two new args to CrossTable() to allow
> you to specify these names explicitly, rather than having them as the
> way they are now, which simply takes the x and y args and does:
> 
>     RowData <- deparse(substitute(x))
>     ColData <- deparse(substitute(y))
> 
> The result is that whatever is passed as the x and y arguments, will be
> used as the titles for the row and column labels as you have noted.
> 
> In the mean time, I am attaching an update to CrossTable (which I have
> not extensively tested yet), that you can source() into R via the
> console. The update has two new args called "RowData" and "ColData"
> which will default to NULL, so as to not impact current default
> behavior. You can then set these as part of your loop by passing the
> index values.
> 
> 
> Using one of the examples in ?CrossTable:
> 
> > CrossTable(infert$education, infert$induced, RowData = "Education",
>             ColData = "Induced")
> 
> 
>   Cell Contents
> |-------------------------|
> |                       N |
> | Chi-square contribution |
> |           N / Row Total |
> |           N / Col Total |
> |         N / Table Total |
> |-------------------------|
> 
> 
> Total Observations in Table:  248
> 
> 
>             | Induced
>   Education |         0 |         1 |         2 | Row Total |
> -------------|-----------|-----------|-----------|-----------|
>      0-5yrs |         4 |         2 |         6 |        12 |
>             |     1.232 |     0.506 |     9.898 |           |
>             |     0.333 |     0.167 |     0.500 |     0.048 |
>             |     0.028 |     0.029 |     0.162 |           |
>             |     0.016 |     0.008 |     0.024 |           |
> -------------|-----------|-----------|-----------|-----------|
>     6-11yrs |        78 |        27 |        15 |       120 |
>             |     1.121 |     1.059 |     0.471 |           |
>             |     0.650 |     0.225 |     0.125 |     0.484 |
>             |     0.545 |     0.397 |     0.405 |           |
>             |     0.315 |     0.109 |     0.060 |           |
> -------------|-----------|-----------|-----------|-----------|
>     12+ yrs |        61 |        39 |        16 |       116 |
>             |     0.518 |     1.627 |     0.099 |           |
>             |     0.526 |     0.336 |     0.138 |     0.468 |
>             |     0.427 |     0.574 |     0.432 |           |
>             |     0.246 |     0.157 |     0.065 |           |
> -------------|-----------|-----------|-----------|-----------|
> Column Total |       143 |        68 |        37 |       248 |
>             |     0.577 |     0.274 |     0.149 |           |
> -------------|-----------|-----------|-----------|-----------|
> 
> 
> Let me know if this works or you find a problem. I will do further
> testing here as soon as time permits and get an update to Greg and Nitin
> to include into gregmisc.
> 



1. Assuming that the names of the data frame, dat, are 
set as desired in the output, then with RowData= and ColData= 
arguments implemented the key portion of the poster's problem 
could be written 

  CrossTable(dat[,i], dat[,j], 
    RowData = names(dat)[i], ColData = names(dat)[j])

2. However, instead of naming the new args as RowData= and
ColData= they might be named consistently with the table
function using the argument named dnn= .  In that case the
above could be shortened to:

   CrossTable(dat[,i], dat[,j], dnn = names(dat)[c(i,j)])

3. Even better would be to allow a data frame argument with
automatic use of the names in that data frame in which
case the example becomes just:

   CrossTable(dat[,c(i,j)])


By the way, here is a solution that can be used even with
the existing version of CrossTable. The first portion just
sets up test data assuming we want to specify columns 1 and
3.  The second portion substitutes the column names into the
expression giving s and then evaluates s in the context of dat.

   library(gmodels)

   dat <- data.frame(a = 1:2, b = 1:2, c = 1:2) 
   i <- 1; j <- 3

   nm <- lapply(names(dat), as.name)
   s <- substitute(CrossTable(coli,colj,chisq=TRUE,format="SPSS"),
      list(coli=nm[[i]], colj=nm[[j]]) )
   eval(s, dat)