row number with by in data.table

Question

I have this easy task to do, imagine I have this table :

   var1
1:    X
2:    X
3:    Y
4:    X
5:    Z
6:    Y

And I want to create a new variable like this :

   var1 nth
1:    X   1
2:    X   2
3:    Y   1
4:    X   3
5:    Z   1
6:    Y   2

In reality I have a lot of variables plenty of lines. So my actual solution is (which works well) :

dt[, flag := 1]
dt[, nth := frank(flag, ties.method = "random"), by = .(var1)]

but this is slow, and I'm not really interested by ranking.(frank has to sort and rank, and I need just a simple thing which is indexing).

I imagined this solution, but doesn't work :

dt[, pos := .I, by = .(var1)]

Do you have any fast and straightforward solution for this task ?

Thank you

Answer 1

dt[, nth := row.names(.SD), by = var1]