But add_count() from another comment seems to do the same with less typing, except it doesn’t change x to a factor.

Okay, tested that in R. Maybe not what you wanted.

data[, x := factor(x)][, N := .N, by = .(x)][order(-N)]

might be closer. If you want unique values of the factor counted in a second column and to order those as well, it would go: by = .(x,y) and order(x,y,-N) as needed.

I can’t be sure this will work the way you want without checking some things but I would do something like:

library(data.table)
setDT(data)
data[, .N, by = .(x)][order(N)]

Maybe order(-N) if you want the sorting reversed, and .(factor(x)) if default factor level order is fine and it behaves.

I’m now selling this as an NFT. Starting price is $45 billion.

I’ve been following many/all of the ones I’ve run across for a couple of months so you can probably start with my Following list. There should be a couple hundred on there.

So excited! That’s good timing with our supposed “less busy” period. I won’t say anything about the date to my huge following, of course, but will check the website every few minutes in early January. Thank you! 😁