Predicting the future with Fourier transforms - page 42

 
Trololo:


I also don't understand the reason (I understand that it only shows the state of the segment). that's why I wrote that it's not that simple.

You're saying that if someone proves that Fourier can't be applied here, are you sure that this person has tried and tried absolutely every possible Fourier application?

I'm not talking about primitive - just pick and place on a chart, for example. There's plenty to try out, and you don't know for sure if you've tried everything.

Only it won't be Fourier anymore.
 
AlexeyFX:

It has been correctly written here that the Fourier transform only applies to periodic functions. But still there are those who want to pull it on forex. They think they can analyse, predict and make money before the spectrum changes. So, it's not the changeability of the spectrum that matters, but the fact that the Fourier decomposition is wrong in non-periodic functions .Take a section of a sine wave that is exactly 1 period long and decompose it by Fourier. You get a single harmonic, as it should be. Take a section of the same sine wave not a multiple of a period and you get a bunch of harmonics that are not in the original signal. That's the whole explanation of the 1st Fourier problem on your fingers.

Sorry, but this is not an explanation of Fourier, but a demonstration of its complete lack of understanding.
 

I have not waited for an answer to my question and I haven't made any pictures yet(((( what's the point of creating a branch because of such nonsense.

maybe someone will take a look and give me some good advice on how to calculate? https://www.mql5.com/ru/forum/108103/page39 down there. thanks.

 
AlexeyFX:


It has been correctly written here that the Fourier transform only applies to periodic functions.

Fourier Rows, Fourier Numbers apply to periodic functions. The Fourier transform applies to any function!
 
alsu: The Fourier transform can be applied to any function!
You can apply it to anything, but what do you get? )))
 
filatura, Len, filatuuruuuu))
 
LeoV:
You can apply it to anything, but what do you get? )))

Spectrum. Approximation.
 
Integer: Spectrum. Approximation.

Agreed, but I thought we were talking about making a profit here....((((
 
alsu:
Rows, Fourier RNGs apply to periodic functions. The Fourier transform applies to any function!


Isn't the transformation a series expansion?

Decompose, add, you get the same thing, works on anything.

 
LeoV:
You can apply it to anything, but what do you get? )))
Just talking about Fourier we somehow mean only numerical discrete PF of price series. We should not forget about (apparently obsolete) way to get results analytically by means of theoretical calculations, where no one prevents to use a normal continuous PF)
Reason: