Create n-way frequency tables using R - r

Question

I need some help to create a n-way frequency table.
I am using the code below:
tab <- table(VAR1,VAR2,VAR3)
finaltab <- ftable(tab,row.vars=c(2,3))
print(finaltab)
VAR1, VAR2 and VAR3 are all factor variables. By doing this, I produce the following table:
But since VAR2 and VAR3 have several categories, I got a lot of lines with "0" and I which to remove those lines to keep in which category of VAR2 only frequencies for the categories of VAR3 that really have frequency values, as follows:
Does anyone knows how to do it, either by subsetting the table I created first or using another function that doesn't return all levels of VAR3 in each VAR2 category but only those which actually have frequencies?

Answer 1

Contingency tables have the same number of rows in every category. If you
remove rows from one category you no longer have a table but a matrix.
t <- structure(c(0L, 0L, 1L, 1L, 1L, 0L, 1L, 1L, 1L, 0L, 0L, 0L, 0L, 1L, 0L, 1L, 1L, 1L), .Dim = c(3L, 2L, 3L), .Dimnames = structure(list(c("A", "B", "C"), c("A", "B"), c("A", "B", "C")), .Names = c("","", "")), class = "table")
> (ft <- ftable(t, row.vars=c(2,3)))
A B C
A A 0 0 1
B 1 1 1
C 0 1 0
B A 1 1 0
B 0 0 0
C 1 1 1
> ft[apply(ft, 1, any), ]
[,1] [,2] [,3]
[1,] 0 0 1
[2,] 1 1 1
[3,] 0 1 0
[4,] 1 1 0
[5,] 1 1 1
The unfortunate effect of subsetting the table is that names are lost. This
can be mitigated to some extent by coercing the table to a matrix before
taking the subset, but the printed output still isn't as pretty as that of a
contengency table.
> as.matrix(ft)[apply(ft, 1, any), ]
_ A B C
A_A 0 0 1
A_B 1 1 1
A_C 0 1 0
B_A 1 1 0
B_C 1 1 1