|
PureBytes Links
Trading Reference Links
|
i spoke to omega tech support last tuesday regarding this issue and they were
supposed to e-mail
to this list a solution to this problem.
i was told not to change the price scale but to edit the first bad tick with the
same value as the last
good tick(continuing so until no bad tick existed), all this while in the edit
tick screen, i used
a 0.5% filter.
today the problem resurfaced again.
unless i have missed it i have neither seen an e-mail nor an answer in their
knowledge base.
this is poor customer support.
bmi is not the problem the software is and gary's posting was very insightful
into what actually been
happening.
on another subject take a look at omga and bmi chart(dbcc).is the market telling
us something?
Hango Winspy wrote:
> **** Date: Tue, 11 Apr 2000 08:24:19 -0600 ****
> Gary wrote
>
> >TS apparently stores intraday prices as offsets from the opening
> >price. They use a 16-bit integer to store these offsets, multiplying
> >the 16-bit integer times the PriceScale set up for that symbol.
> >Thus, TS can handle a maximum offset from the open of 2^16/2 *
> >PriceScale. If your ND / NQ / NDX / etc is set at the default
> >PriceScale of 1/100, that maximum offset is 327.68. If you exceed
> >that maximum offset, the 16-bit integer overflows and you get a
> >sudden jump of 2^16 * PriceScale = 655.36.
>
> >The open for ND0M yesterday was 4350, so the minimum price TS could
> >store with a 1/100 PriceScale was 4350 - 327.68 = 4022.32. The price
> >hit went below 4022 at 4:00 yesterday. Any ticks below 4022.32 would
> >have been stored with a value 655.36 too high.
>
> Gary and All,
>
> Thanks Gary, I accept your explanation and theory. I'll change my
> PriceScale as you recommended. It's much easier and painless, also it
> should work as you stated.
>
> Please disregard my previous post about how I saved the data earlier this
> morning. As I stated, I did not have any theory to back up my way of saving
> the "clean" data. I don't know why, but it did work few times in the past.
> This was probably pure luck, just like a blind cat caught a dead mouse <G>.
> It's "luck", "luck, ... I guess, nothing more and nothing less. Every time
> when I failed to save the clean data, I thought it was because I didn't save
> them in time. Before I read your explanation, I thought I had some
> significant discovery. Well, I surrender!
>
> Thanks again, Gary
>
> HW
|