How do I make big.mark apply to more than just the first column? - r

Question

Here's what I'm getting:
> panderOptions('big.mark', ',')
> foo <- rbind(cancer, cancer); for(i in 1:8) foo <- rbind(foo, foo)
> pander(table(foo$ph.karno, foo$pat.karno))
-----------------------------------------------------
30 40 50 60 70 80 90 100
--------- ----- ---- ---- ---- ---- ----- ----- -----
**50** 0 512 512 512 512 512 0 0
**60** 0 512 512 2560 4608 1536 0 0
**70** 1,024 0 1024 4608 3072 3072 1536 1536
**80** 0 0 0 6144 6656 6656 10240 4608
**90** 0 0 0 1536 5120 10752 12288 7680
**100** 0 0 0 0 512 3584 6656 4096
-----------------------------------------------------
I would like the comma delimiter to show up in the other columns too. How do I do that?

Answer 1

Best result I got (with t being your table call result):
> pander(format(t,big.mark=','))
----------------------------------------------------------
30 40 50 60 70 80 90 100
--------- ----- ---- ----- ----- ----- ------ ------ -----
**50** 0 512 512 512 512 512 0 0
**60** 0 512 512 2,560 4,608 1,536 0 0
**70** 1,024 0 1,024 4,608 3,072 3,072 1,536 1,536
**80** 0 0 0 6,144 6,656 6,656 10,240 4,608
**90** 0 0 0 1,536 5,120 10,752 12,288 7,680
**100** 0 0 0 0 512 3,584 6,656 4,096
----------------------------------------------------------
I assume this is a bug in pander.table.return but I did not dig enough to get the root cause.
Edit: I found the reason, line 283 of the function there's a for loop calling sapply on each col, but once the first column has been processed, the full array of the table is converted to char as we bring in chars from the output of format.
Then the subsequent calls to format can't format the number as they've been coerced to char.
Code from pander.table.return giving this behavior:
for (j in 1:ncol(t)) {
temp.t[, j] <- sapply(temp.t[, j], format, trim = TRUE,
digits = digits[j], decimal.mark = decimal.mark,
big.mark = big.mark)
}