ERUUSD spectra - is this proof of non-stationarity? - page 3
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
A non-stationary signal (flow, series, process) does not and cannot have a SPECTRE.
In fact.
By definition - both.
that's like saying there's no time axis in the quotes.
What's your problem? Spectrophobia?
Just because people on the forum don't know (including you) what to do with a spectrum doesn't mean they should be shielded from spectral research
show them the right way
>> you don't have one.
You're just making your brain laugh with your inter-banking.
A non-stationary signal (flow, series, process) does not and cannot have a SPECTRE.
At all.
By definition - both.
As far as FFT is concerned, I agree. The attached graphs are obtained with a digital method generator that does not use FFT, but uses Berg's approach and it is not the spectrum that is estimated, but the spectral power. In that sense there is always a spectrum at a particular point in time. We see a spectrum on the 0 -480 bar segment, but another on the two segments that make up the first.
There are no sustained frequencies, so spectral analysis really doesn't work.
There are no sustained frequencies, so spectral analysis really doesn't work.
Kravchuk claims that it works. I have been looking at the FIR filters branch. I did not find an answer to my question. It's not just the stability of the frequencies, but also the window in which we are trying to define them. On my graphs it is 240 and a multiple of 480. Maybe a different window will give a better result.
Kravchuk claims that it works. Watched the FIR filter branch. Did not find the answer to my question. It is not only stability of frequencies, but also the window in which we are trying to define them. On my graphs it is 240 and a multiple of 480. Maybe a different window will give a better result.
You look at the year when Kravchuk wrote these articles. And before he wrote them, he was still working on it before anyone knew it. Since then the market has long since changed as many people found out about it, and when everyone knows about it, it doesn't work......
There are no stable frequencies, so spectral analysis really doesn't work.
Why do you need them to be stable? The frequencies are there! Take it and use it!
Very interesting picture. If it works for you, give me a forecast on GBPUSD when it is going to reverse there, I'm sick of the flat.
Why do you need them to be stable? The frequencies are there! Take it and use it!
In the past. as you can see from my graphs H1 -240- -480 are some frequencies, and in the next section 0 - -240 are others, and what will be after the ban #0?
You look at the year Kravchuk wrote these articles. And before he wrote them, he was still working on it before anyone knew it. Since then, the market has long since changed as many people found out about it, and when everyone knows about it, it doesn't work......
It's not about Kravchuk. Mashka is a filter with a coefficient. 1/period. In the proposed graphs we get the same mashka, but the coefficients are different, several dozen. Which is better?
It's not about Kravchuk. Mashka is a filter with a coefficient of 1/period. In the proposed graphs we get the same mashka, but the coefficients are different, several dozen. Which is better?
What do you mean by better?