How to count unique combinations from a data table in R?

Question

I have a data table with three columns. The first two are a collection of the data points (categorical data that can be either A, B, or C). The third column is a concatenation o

Answer 1

dt1[,paste(sort(c(CAT1,CAT2)),collapse=" & "),by=1:nrow(dt1)][,table(V1)]

Answer 2

You can also do something like this- Note- As mentioned by @chinsoon12, we can use pmin & pmax

 > setDT(dt1)[,list(Count=.N) ,paste(pmin(CAT1, CAT2), pmax(CAT1, CAT2), sep=' & ')]
   paste Count
1: a & a     1
2: b & b     2
3: c & c     2
4: a & b     2
5: a & c     3

Answer 3

I'm no good with data.table, so here's my answer with a data.frame:

Just sort the two CATs before pasting, making sure they're always in the same order.

 dt1$merged<-apply(dt1,1,function(x) paste(sort(x),collapse=" & "))

I'm sure there's a faster way to do in with data.table, but I'm not sure how. A naive sort added to your code came up with an error...