Comment by slrz

I think you make the Erlang definition look unnecessarily awful.

fac(0) -> 1;

fac(N) -> N * fac(N-1).

Or is there a particular advantage to write it using guards?