Re: Odd problem in trying to write an indicator

["Owen Davies" <owen@xxxxxxxxxxxxx>]

Thanks for your analysis.  For some reason,
I wasn't getting the divide-by-zero messages.
Probably a quirk of the data I was using that the
occasion didn't arise.  Of course, this leaves me
wondering how a problem that did not visibly
arise could have thrown the calculations off,
even if it would have done so on other data.

I confess, I'm unsure how a DBZ could occur,
as the Bollinger top and bottom are not likely
to be identical after the first bar or two, and I
wouldn't have expected those to affect things
later on.  One of the things I've always liked
about Metastock is that it handles DBZ errors
much more gracefully than Omega products.
It's clearly something to iron out, though.

No doubt you're aware that with an oscillator
based on Bollinger Bands, the reading can
go above the usual bounds of 0 and 100.
A reading of -200 is pushing it a bit, but maybe
on a sharp breakout from a narrow range?

Anyway, thanks again.  I'll keep at it.

Best regards,

Owen Davies