Can anyone translate French to explain the Goertzel Amp, Period & Phase Algo

Bard · 2016-11-01T18:03:56+00:00

Hi. Can anyone please explain this indicator as my French is limited. See: http://hk-lisse.over-blog.com and the Blog Post on the 14th May 2009 (bottom 1/3rd of the page) The original French description is: “Dans le dernier article sur l’algorithme de Goertzel, le programme permettait de retrouver les 2 périodes des cycles qui avaient la meilleure réponse. Je l’ai complèté et corrigé pour qu’il sorte également l’amplitude correcte du cycle ainsi que la phase : cela afin de pouvoir redessiner la courbe. Dans l’exemple ci-dessous, le signal (en noir) est une sinusoïde de période 11 ajoutée à une sinusoïde de période 19 (d’amplitude 2) avec un décalage de phase (90), le tout avec une pincée de bruit.. Les 2 courbes (rouge et bleue) sont reconstruites par le programme, il reste à les superposer pour avoir le signal sans bruit.” Which I have “translated” as: “In the last article on the Goertzel algorithm, the program allowed to find the 2 periods of cycles that had the best answer/fit. I completed and corrected it to also ensure the correct amplitude of the cycle and the phase in order to redraw the curve. In the example below, the signal (black) is an 11 period sinewave added to a 19 period sinewave (amplitude 2) with a phase shift (of 90 degrees), all with only a tiny bit of noise. The two curves (red and blue) are reconstructed by the program, it remains to superimpose to have the signal without noise.” But this translation doesn’t really help me fully understand the indicator and how to use it or what to set the value of K to? I am not clear in how this algorithm indicator helps predict the turning points in a market although I am familiar with John Ehler’s MESA Cycle/Phase concepts. I guess a good translation would be a start. It seems John Meyer gets good trading system results with high % of winning trades with good profit factors using the Goertzel Algorithm. Any help would be great, Thanks in advance, ProRealTime Code: //k = the frequency already found// // sous prgm goertzel biss // a=barindex sig=sin(360*a/11)+2*sin(360*a/19+90) b=SQRT(barindex*(b+1)) b=abs((b-ROUND(b))*4) c=b-1 sig=sig+c n=200 pr=sig alpha=2*cos(360*1/k) q1=0 q2=0 for i=n downto 0//fenêtre de 200 barres q3=pr[i]+alpha*q1-q2 q2=q1 q1=q3 next amp=sqrt(square(q1)+square(q2)-alpha*q1*q2) real=q1-(alpha*q2)/2 imag=q2*sin(360*1/k) phas=atan(imag/real) if real =0 and imag<0 then phas=phas+360 endif return (2*amp/n)*sin(360/k+phas[1]+90),sig

Viewing 4 posts - 16 through 19 (of 19 total)

← 1 2

Author

Posts
04/08/2019 at 8:26 AM #95672 quote

Nicolas
Keymaster

Legend

A lot of John Ehlers indicators are available here: https://www.prorealcode.com/tag/john-ehlers/

Is this one useful for your query? @Finning
Ehlers’s Adaptive Cyber Cycle

04/08/2019 at 12:21 PM #95695 quote

Finning
Participant

Veteran

Hi Nicolas,

I have had a look through the Ehlers tags, and I couldn’t find what I was looking for.

I am looking for an indicator to find the dominant cycle of the market.

The Cyber Cycle doesn’t do what I’m looking for either – however apparently it does use the dominant cycle in the code – just don’t exactly know to isolate that part of the code.

I did see this below request in the forums – https://www.prorealcode.com/topic/convert-autocorrelation-periodogram-cci-from-easy-language-to-prorealtime/

This would have done the job perfectly – however the Easy Language code needed arrays to work – which PRT doesn’t support.

What I’m ultimately trying to achieve is a measurement that I can then average to describe the dominant cycle of the market – and a Dot Plot of the dominant cycle – just as on the HK-Lisse page – would be great.

The reason I’m looking for this dominant cycle is to primarily tune indicator lengths to market phase.

Seems a bit unfair to ask you to find the missing pieces (a MESA based dominant cycle finder), but if you could – and post it to the library for all, that would be much appreciated.

Finning.

PS – have also seen this – https://www.prorealcode.com/topic/dominat-cycle-calculation-by-band-pass-corssing-over/

I don’t believe that this is the “original” way to find the dominant cycle. I have got this method to work, and get results that are reasonably correlated to my empirical cycle measurements – however a dot plot or autocorrelation periodogram of dominant cycle information is much more useful in revealing additional cycle information.

04/09/2019 at 8:13 AM #95782 quote

Nicolas
Keymaster

Legend

You’ll find attached to the post the ITF file for the Goertzel cycle detection algorithm. When you import the file you’ll get all necessary sub-functions imported at once.

Bear in mind that the indicator will find cycle period from 5 to 25, because it was an experiment to find the correct one that HK-Lisse introduces in his article, he was using a simple fake signal (the first red curve = no noise curve) and then he made another test with another fake signal but with more noises (the green curve).

So in order to get a dominant cycle of the price data serie, you must make some changes in the code (increase the range of cycle period for instance) and also inject the price serie in a detrended format, otherwise you’ll get false results obviously.

Bard thanked this post

prgm-goertzel.itf

04/09/2019 at 12:50 PM #95814 quote

Finning
Participant

Veteran

Hi Nicolas,

thank you very much for that! I will test shortly, and let you know how it goes.

Regards,

Finning
Author

Posts