I’m trying to create an index column based in multiple variables with a behavior like rleidv in
data.table, as an example taking just a few cases of mtcars dataset

library(collapse)
mtcars |> fsubset(cyl == 4) |>
  fselect(cyl, gear, carb) |>
  roworder(cyl, gear, carb) |> 
  fgroup_by(cyl, gear) |> 
  ftransform(id = groupid(carb))

The output is as below

               cyl gear carb id
Toyota Corona    4    3    1  1
Datsun 710       4    4    1  1
Fiat 128         4    4    1  1
Toyota Corolla   4    4    1  1
Fiat X1-9        4    4    1  1
Merc 240D        4    4    2  2
Merc 230         4    4    2  2
Honda Civic      4    4    2  2
Volvo 142E       4    4    2  2
Porsche 914-2    4    5    2  2
Lotus Europa     4    5    2  2

What I want to achieve? For each combination of cyl and gear I want to make an index of the carb variable starting always in 1 , so the output should be like

               cyl gear carb id
Toyota Corona    4    3    1  1
Datsun 710       4    4    1  1
Fiat 128         4    4    1  1
Toyota Corolla   4    4    1  1
Fiat X1-9        4    4    1  1
Merc 240D        4    4    2  2
Merc 230         4    4    2  2
Honda Civic      4    4    2  2
Volvo 142E       4    4    2  2
Porsche 914-2    4    5    2  1
Lotus Europa     4    5    2  1

is there a way to do it with collapse?