|
PureBytes Links
Trading Reference Links
|
Well, my fought was that since the function is cyclic, you are doing the
same thing every 360 bars 180 of them are the opposite of the other 180.
Besides xaverage does not accumulate by design, so it did not feel like a
good test.
> -----Original Message-----
> From: pierre.orphelin [mailto:pierre.orphelin@xxxxxxxxxxxxxx]
> Sent: den 30 juli 2001 15:58
> To: omega-list@xxxxxxxxxx
> Subject: RE: TradeStation Precision - Summary
>
>
>
>
> > -----Message d'origine-----
> > De : Bengtsson, Mats [mailto:mats.bengtsson@xxxxxxxx]
> > Envoyé : lundi 30 juillet 2001 13:53
> > À : MikeSuesserott@xxxxxxxxxxx; pierre.orphelin@xxxxxxxxxxxxxx;
> > omega-list@xxxxxxxxxx Objet : RE: TradeStation Precision - Summary
> >
> >
> > I would suggest trying to avoid sin/cosine or any other cyclic
> > function. Best thing would be to use real data (both
> operating on the
> > same data series, like djia for example) whicheverybody could get
> > their hands on of they wanted to repeat the experiment, and
> then use
> > something that is on the line of what everybody tries to do
> in one way
> > or the other (signal to noice filtering in one way or the other).
> >
>
> The sine cosine function produces a synthetic variable price
> that is correct up to the 7th digit , and allowed me to
> highlight the propagation error effect.
>
> Using price data to test the difference with TS implies that
> I must use an other program to have a benchmark, what I have not.
>
>
> PO
>
This message contains information that may be privileged or confidential and is the property of the Cap Gemini Ernst & Young Group. It is intended only for the person to whom it is addressed. If you are not the intended recipient, you are not authorized to read, print, retain, copy, disseminate, distribute, or use this message or any part thereof. If you receive this message in error, please notify the sender immediately and delete all copies of this message.
|