What is the general formula for a rate in epidemiology and why are rates used instead of absolute counts?