I've modified the code as I mentioned in a previous reply. If you want to see
it, let me know. It will most likely be lower to calculate than a formula, but
if you decide you want to see data for periods of different lengths, it will
be a lot easier to generate the formula <g.