|
PureBytes Links
Trading Reference Links
|
Exchange data will always be dirty because it is created by humans
- who make errors.
1. Either the data is to dirty that you could never trade that market
in real-time anyway then there is no point of cleaning the data and
developing some system. The solution then is to get better data
(real-time).
2. Or the the data is just sometimes dirty then you have to identify
the dirt mathematically and remove it in a consistent manner
mathematically. The way I have approached this is by using only
filtered data for systems. Basically I use the result of this
non-smoothing-dirt-filter as the input for the systems in real-time
and historically. This way I know that the filtering method is
consistent.
Gerrit Jacobsen
> At 2:09 AM -0500 1/10/99, Tony Parker wrote:
>
> >I'm looking for a (free or subscribed) source of "Cleaned" Downloadable
> >Historical Intra-Day Data. I'm also looking for a source of "delayed"
> >"Cleaned" Intraday Data that is in a downloadable format much like the Omega
> >Research Intraday Data but is "Cleaned" of all the Bad Ticks that I find in
> >the Omega data.
> >
> >The trouble with the so-called "free" downloadable Intraday data found on
> >the Omega Research Web-Site is that it's just as "Dirty" and as full of bad
> >ticks as you get from any of the "live" RT (Signal, BMI, etc.) sources.
> >
>
> <snip>
>
> >
> >I spend at least two hours every night cleaning the data manually...and I
> >still get inconsistent results. There has to be another way!
> >
> >Ideas anyone?
> >
>
> Make me wonder why you think you need clean data. If the data you get
> real-time is "dirty" (and it is), then what good is a trading system that
> only works on clean data.
>
> Wouldn't it be more appropriate to have one that works on dirty data since
> that is what is available in real-time?
>
> Bob Fulks
>
>
|