Fast Fourier Transform - Cycle Extraction - page 100

New comment
 

UltraBanker, you keep talking about "we"...who are you referring to?

 
Snowski:
UltraBanker, you keep talking about "we"...who are you referring to?

Look at my homepage in my profile here on forum, please.

Thank you.

 

Ultrabanker, with regards to point 5a and 5b....why do think that?

My opinion is different than yours. I personally feel that extracting a cycle in a forex time series is far more reliable than equities because there is inherent symmetry. Unlike equity indexes were there is an overwhelming long bias with far more money behind long positions than short.

I am sorry say that while you obviously spent plenty of time writing your post, in the end you really did not say anything that added any value to this thread. I hope you do return in a couple of years when your team is ready to prove the skeptics like myself wrong.

 
hughesfleming:
Ultrabanker, with regards to point 5a and 5b....why do think that?

My opinion is different than yours. I personally feel that extracting a cycle in a forex time series is far more reliable than equities because there is inherent symmetry because of being made up of two equal components. Unlike equity indexes were there is an overwhelming long bias with far more money behind long positions than short.

I am sorry say that while you obviously spent plenty of time writing your post, in the end you really did not say anything that added any value to this thread. I hope you do return in a couple of years when your team is ready to prove the skeptics like myself wrong.

Thank you for Your suggestion about our results in the future.

Please, do not take my answers as too arrogant: I am just keeping secret the scientific results and commercial secrets of our team.

As for (5a) and (5b):

Preface: before starting the development of the spectrum trading system, please ask yourself why thousands of engineers, being familiar with Fourier interpolation theory - why they are not billionaries? - if they can EASILY extract cycleS (many cycles) from price stream?

Personally I have spent 10+ years in the scientific research of stupid digital sound, and found alternatives. Why there should be alternatives to Kotelnikov-Shannon sampling theory ? - THAT is THE question.

I have found alternatives long ago and they are VERY limited. I am not trying to hide here the right way to the solution. While knowing the right way I am simply warning other traders here about VERY SERIOUS DIFFICULTIES on this way: both in sampling theory (and Fourier's methods) and in the practical implementation of any other alternatives. Do not be surprised when you will see that ANY spectral method, described in books available in English, will sometimes work and more often will not.

 

Ultrabanker.....regardless of how the cycles are calculated while looking at the past, the problem is always forecasting the future. Oh...and the reason most engineers are not billionaires is because the vast majority are not traders.

Don't get me wrong, I am very interested in new approaches to this old problem. Please continue to post once you have your project off the ground.

Alex

 
hughesfleming:
Ultrabanker.....regardless of how the cycles are calculated while looking at the past, the problem is always forecasting the future. Oh...and the reason most engineers are not billionaires is because the vast majority are not traders.

Don't get me wrong, I am very interested in new approaches to this old problem. Please continue to post once you have your project off the ground.

Alex

Let me repeat by other words: you can obtain the "spectrum" of the "signal" by incorrect Fourier method or other way. If the signal is stable within the longest period in spectrum, then you can easily extrapolate the same spectrum into the future. That is the way the "Extrapolator" from MQL codebase works.

But if your "signal" has a spectrum which is changing over time - then you need a "dance" of the spectrum - this is "spectrum of spectrum" or somewhere called "Cepstrum" - a dynamic representation of spectra's changes over time. Once you have got a cepstrum, you can extrapolate it to get future spectrum, then you can extrapolate spectrum for the future time, and then you can extrapolate the "signal" - the price.

The faster the spectrum changes the more difficult to get the "cepstrum".

That is what usually the speech recognition engineers are dealing with. This is much more difficult task for engineers, than simple spectra extraction. That is the main difficulty in what the Forex is a champion.

Even if you got a cepstrum, and a spectrum, you cannot be sure that this your so-called "spectrum estimation" - a "probable suggestion" in simple words - is correct. Therefore you must periodically return to cepstrums, obtained and refused earlier by your program (by some reasons) - in order to correct your trading system. As I understand Simons and RenTec solved this problem by using VERY exotic approach - they hired as a chief scientist one on the main specialists in so-called "returning functions". And they always optimize obtained cepstums ans spectrums as to the new turns of the market. They never sure that the spectra they use at the moment for predicting the futue - is correct.

You see? No place for amateurs. This is a very serious math job.

 

Greetings all,

Given Ultrabanker's thorough posts (thank you btw), I feel, somewhat, inadequate with the request that follows. But, for those who are still entertaining the usefulness of cycle extraction the poor-man's way I would like to request a modification to the Spectr indicator.

As it stands, the Spectr indicator has a meter function in it that qualitatively measures the relevance of the cycle that is in play based on amplitude; I would like to ask whether it is possible to introduce a feature that takes into account this information and automates the generation of the composite cycle based on amplitude of the cycle. For that, I believe, the following variables would be introduced.

1) Boolean Auto_Composite_Cycle True

2) int Number_Of_Cycles 2

3) int Amplitude_Threshold 20

-The first variable is simply a Boolean that turns the feature on or off and is subordinate to the existing Boolean of CompositeCycle

-The second variable dictates the maximum number of cycles to include in the composite cycle

-The third variable indicates the minimum amplitude threshold that the cycle has to pass in order to be considered.

Thanks,

Pip

 
Pip:
Greetings all,

Given Ultrabanker's thorough posts (thank you btw), I feel, somewhat, inadequate with the request that follows. But, for those who are still entertaining the usefulness of cycle extraction the poor-man's way I would like to request a modification to the Spectr indicator.

As it stands, the Spectr indicator has a meter function in it that qualitatively measures the relevance of the cycle that is in play based on amplitude; I would like to ask whether it is possible to introduce a feature that takes into account this information and automates the generation of the composite cycle based on amplitude of the cycle. For that, I believe, the following variables would be introduced.

1) Boolean Auto_Composite_Cycle True

2) int Number_Of_Cycles 2

3) int Amplitude_Threshold 20

-The first variable is simply a Boolean that turns the feature on or off and is subordinate to the existing Boolean of CompositeCycle

-The second variable dictates the maximum number of cycles to include in the composite cycle

-The third variable indicates the minimum amplitude threshold that the cycle has to pass in order to be considered.

Thanks,

Pip

Point 1. : there is already a parameter that can turn the composite on or off - the effect would be the same as making it "automatic" or "not automatic"

Point 2. : you can chose any cycles that are making the "composite" cycle

Point 3. : "spectrometr" is not honoring the amplitude values the way how Fourrier transform does that. Have no idea why the original author did so (my guess is to make all the cycles phases equally visible). Also, excluding amplitudes on one bat could be highly misleading : for example, the amplitude of the most significant cycle could be exclude just because at some bar it has a value of 0

 
sebastianK:
Anybody has this type of FFT for tradestation?

FXCodeBase.COM: Forex Chart Indicators and Development • View topic - Fast Fourier Transformation indicator

You want this one?

 
gezis:
FXCodeBase.COM: Forex Chart Indicators and Development • View topic - Fast Fourier Transformation indicator You want this one?

gezis

Thank you, but that is for trading station and I need a version for tradestation

1...93949596979899100101102103104105106107...112
New comment
Reason: