[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[EquisMetaStock Group] Fourier Transform



PureBytes Links

Trading Reference Links

You can buy back issues of TASC, or the TASC reference disc. You can 
also read John Ehler's books. 

You can subscribe to Roy's newsletter and the simple version of some 
of this maybe get mentioned in there. Or you can create your own 
versions. 

The Fisher Transform in MS function language:

pr:=(H+L)/2;
len:=10;
maxh:=HHV(pr,len);
minl:=LLV(pr,len); val1:=.33*2*((pr-minl)/(maxh-minl)-.5)+.67*PREV;
value1:=If(val1>.99,.999,If(val1<-.99,-.999,val1));
fish:=.5*Log((1+value1)/(1-value1))+.5*PREV;

fish;
{Ref(fish,-1);}
10*ROC(fish,1,$)

I'm talking about the inverse of the above. The simplified 
mathematical reference can be seen here.

http://cswww.essex.ac.uk/PLANET/summer-school-02/materials-
revised/SummerSchoolSlides/Part2/img8.htm

You can read about Laguerre here:

http://www-gap.dcs.st-and.ac.uk/~history/Mathematicians/Laguerre.html
and then here:

http://mathworld.wolfram.com/LaguerreDifferentialEquation.html

It's a bit more complex but very interesting when applied to several 
types of standard indicators. 

When you actually convert a lot of the mathematical swahili to 
english, it's really pretty simple and effective for smoothing and 
noise elimination.

You can't buy David's stuff, but some of it has been published in 
Futures and TASC. I've taken what he's published and created some 
more of the indicators he shows on his website. 

Have fun! 





 



--- In equismetastock@xxxxxxxxxxxxxxx, "James Palmer" 
<steinington@xxxx> wrote:
> superfragalist,
> 
> thanks for the references, just out of curiosity, what is the Fisher
> Transform?
> ----- Original Message -----
> From: "superfragalist" <no_reply@xxxxxxxxxxxxxxx>
> To: <equismetastock@xxxxxxxxxxxxxxx>
> Sent: Thursday, September 16, 2004 11:01 AM
> Subject: [EquisMetaStock Group] Fourier Transform
> 
> 
> >
> > John Ehlers and David Sepiashvili are probably the most 
experienced
> > adapters of engineering mathematics to the market. Neither one 
does
> > much with Fourier.
> >
> > I use a lot of DSP mathematics to create indicators and such. You 
can
> > read Ehler's book "Cybernetic Analysis for Stocks and Futures :
> > Cutting-Edge DSP Technology to Improve Your Trading"
> >
> > and look at David's webstie.
> >
> > http://www.alticom.com/indicators/overview.html
> >
> > for a up-todate look at where the edge of DSP applications are. No
> > sense in starting from scratch.
> >
> > I've had good success with the inverse Fisher Transform and 
excellent
> > success with Laguerre mathematics.
> >
> > In Roy Larsen's newsletter, www.metastocktips.co.nz he has a 
series
> > of articles starting with Ehlers Distant Coefficient Filter and
> > ending up with a combination filter that I think you might find 
very
> > interesting that uses some cutting edge DSP filtering.
> >
> > Roy's newsletter is a great read and well worth the $99 
subscrition.
> >
> > I think in one of his latter issues this year, he might be
> > considering some Laguerre code which is very rare and very
> > interesting also.
> >
> > None of this creates anything that is earth shattering, but it 
can be
> > used to create a few unique applications that can improve trading
> > results.
> >
> >
> >
> >
> >
> >
> >
> > --- In equismetastock@xxxxxxxxxxxxxxx, "James Palmer"
> > <steinington@xxxx> wrote:
> > > I'm curious if anyone has worked with the Raw Fourier Transform.
> > >
> > > I'm getting results orders of magnitude different from the
> > interpreted.
> > >
> > >
> > > For instance on the S&P500 on the interpreted  I get three 
cycles.
> > 256, 683,
> > > 85.
> > > On the raw for amplitude I get 18.73, 30.13, 38.36.
> > > for spectrum I get 12.13, 31.37, 50.85.
> > > This is using detrend and data sample of 512 days.
> > >
> > > Can anyone explain what the intrpreted is giving me and why 
it's so
> > > different?
> > >
> > >  I assume it's giving me the amplitude.
> > > Also on the raw can you match the amplitude to the spectrum so
> > that, for
> > > instance, the 12.13 day cycle has an ampltude of 18.73 points?
> > >
> > > Much obliged.  I've been planning to code my own FFT but I 
thought
> > I would
> > > first do some experimentation with metastock's built in but 
there
> > is minimal
> > > documentation.
> >
> >
> >
> >
> >
> >
> > Yahoo! Groups Links
> >
> >
> >
> >
> >
> >
> >




------------------------ Yahoo! Groups Sponsor --------------------~--> 
Make a clean sweep of pop-up ads. Yahoo! Companion Toolbar.
Now with Pop-Up Blocker. Get it for free!
http://us.click.yahoo.com/L5YrjA/eSIIAA/yQLSAA/BefplB/TM
--------------------------------------------------------------------~-> 

 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/equismetastock/

<*> To unsubscribe from this group, send an email to:
    equismetastock-unsubscribe@xxxxxxxxxxxxxxx

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/