Help with Fourier - page 13
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Well, all right!!! I lied. Just leave me alone.
==========
But I'll post a picture. I promised you a year and a half ago.
Fourier is not applicable in forex.
Fourier is not applicable in forex.
There is no doubt about it!!! Also all standard indicators are not applicable.
No doubt about it!!! Also, all the standard indicators do not apply.
What is applicable?
>> What's applicable?
It's a half-joke. I'm applying Fourier.
Help who knows:
When decomposing a series using the Fourier method, significant outliers occur at the ends of the period. How to achieve a good approximation at the ends, at least comparable (in terms of RMS) to the approximation inside the period?
Example - in the ettachment
Help who knows:
When decomposing a series using the Fourier method, significant outliers occur at the ends of the period. How to achieve a good approximation at the ends, at least comparable (in terms of RMS) to the approximation inside the period?
Not surprising, since at 0 and 2 * PI the inverse Fourier transform gives the value of the 0th harmonic, i.e. the arithmetic average of the entire period.
Yes. I'm sure. That's why it's called a random process. Otherwise, it should be a quasi-random process, etc.
Sergey, I respect your intelligence and inquisitiveness of mind. I always read your posts with the deepest interest. Your knowledge of mathematics greatly exceeds mine, but let me still disagree with you reasonably:
Red is the frequency in bars by X and the number of hits at a given frequency by Y, below is the amplitude component (an abstract normalised value). The white marker is at a peak frequency of 1500 min (25 hours and 100 bars). The phase is equal to pi.
The calculation depth is 11500 bars (all that was loaded into the terminal at the moment. High frequency components have been filtered out. Pictures from different TFs correlate with each other.
Can we get back to the question of market randomness?