Sort a contingency table without changing table class

Question

I want to reorder factors of cont.table

test.a <- c(rep(1,20),rep(0,40))
test.b <- c(rep(1,25),rep(0,35))

cont.table <- addmargins(table(test.a, test.b))

    test.b
test.a  0  1 Sum
   0   35  5  40
   1    0 20  20
   Sum 35 25  60

I want to be able to order the factors 0 and 1. The result I want is this

     1  0 Sum
1   20  0  20
0    5 35  40
Sum 25 35  60

I did it like this, but I lose the class Table, which is needed for me

> tbl <- as.data.frame.matrix(addmargins(table(test.a, test.b)))
> tbl2 <- cbind(tbl[2],tbl[1],tbl[3])
> tblfinal <- rbind(tbl2[2,],tbl2[1,],tbl2[3,])

> as.table(tblfinal)
Error in as.table.default(tblfinal) : cannot coerce to a table

Is there a possible way? the simpler the better

Answer 1

Make your test.x objects factors with a defined order, then the tables etc will be sorted appropriately. E.g.:

test.a <- factor(test.a,levels=c(1,0))
test.b <- factor(test.b,levels=c(1,0))
addmargins(table(test.a,test.b))

#      test.b
#test.a  1  0 Sum
#   1   20  0  20
#   0    5 35  40
#   Sum 25 35  60

Answer 2

You can just reorder it by the col/row names:

cust_name <- c('1', '0', 'Sum')
cont.table[cust_name, cust_name]

      test.b
test.a  1  0 Sum
   1   20  0  20
   0    5 35  40
   Sum 25 35  60