R中基于另一列的组合

Question

I have not found a question about this precise task, so I'd like to ask how to reach this result in R. I have this dataframe in R.

or<-c("1","1","1","2","2")
pr<-c("a","b","c","w","x")
c<-data.frame(or,pr)

Here the result:

 c
       or pr
    1   1  a
    2   1  b
    3   1  c
    4   2  w
    5   2  x

I'd like to have the -not double, ie not aa,bb,cc,xx,ww- combinations of the column pr , based on the column or . The result is something like:

d
first second
a     b
b     a
c     b
b     c
c     a
a     c
w     x
x     w

The first 6 rows are the combinations of the pr occurrencies with or=1 , and the 7th and 8th row are the combinations of row with or=2 .

I've found the function

expand.grid(c$pr)

But it does not work, above all it does not put the combinations in the frame of 2 column I need.

Answer 1

You can use CJ from data.table with column or as the group variable:

library(data.table)
setDT(c)[, CJ(pr, pr), or][V1 != V2, .(first = V1, second = V2)]

#   first second
#1:     a      b
#2:     a      c
#3:     b      a
#4:     b      c
#5:     c      a
#6:     c      b
#7:     w      x
#8:     x      w

Update :

If you prefer a base R solution, you can use tapply() with expand.grid() :

setNames(
    do.call(rbind, c(make.row.names = F, 
        tapply(c$pr, c$or, FUN = function(v) subset(expand.grid(v, v), Var1 != Var2)))), 
    c('first', 'second'))

#  first second
#1     b      a
#2     c      a
#3     a      b
#4     c      b
#5     a      c
#6     b      c
#7     x      w
#8     w      x