Predicting the future with Fourier transforms - page 19

 
m_a_sim писал(а) >>

Ask Fourier.

loki... If at a given moment in the market the price fluctuations have a (quasi) periodic nature, on the basis of this period we should choose the period of averaging of the machine.

Another variant (not related to the Fourier).

If the trade is conducted, for example, on H1, then perform the optimization for the last week in the tester (or another convenient window period) and get the optimal value of the wavelet period.

Then shift the weekly window by 1 day and optimize it again, etc. This will result in a set of optimal period values for each day.

If these points lie on some more or less smooth curve, we can predict what the averaging period of the waveform for tomorrow should be.

In other words, optimization in the tester can be used as a mechanism for adapting the Expert Advisor's algorithm to the current market conditions - on-the-fly optimization.

But of course, it will not help if the price jumps.

 
'FFT Fast Fourier Transform Functions Library' - link to the Fourier Transform Library
 
diakin >> :

Lolky... If at a given moment in the market price fluctuations have a (quasi) periodic nature, based on this period one should choose an averaging period.

Another variant (not related to Fourier).

If the trade is conducted, for example, on H1, then perform the optimization for the last week in the tester (or another convenient window period) and get the optimal value of the wavelet period.

Then shift the weekly window by 1 day and optimize it again, etc. This will result in a set of optimal period values for each day.

If these points lie on some more or less smooth curve, we can predict what the averaging period of the waveform for tomorrow should be.

In other words, optimization in the tester can be used as a mechanism for adapting the Expert Advisor's algorithm to the current market conditions - on-the-fly optimization.

Of course, this will not help in the case of abrupt price leaps.

that's an interesting thought. Does it work, have you tested it?

 
not tested. I haven't gotten to the optimisation in the tester yet ;)
 
m_keeper

or whoever has figured out this inductor,

Please help me to understand how to use your indicator to
- Determine the moment of a significant phase change (phase change of the main oscillation, possibly without a change in the main frequency)
- and/or a significant change in the oscillation frequency

- Determine for the current bar the fraction in the total cycle of the fundamental frequency, i.e. from the moment when main frequency becomes the current bar, this frequency has passed, for example, 30 or 120% of its period.

Thank you.

 
timbo >> :

A Kalman filter would be much more interesting to watch.

You're falling for some bullshit... It's a bit of a story, you're right... The picture is nothing but "we've seen it"... And you can call all kinds of crap kai-kai, or even non-normative - the essence does not change from this....IMHO - it's a fake, and Prival is just another empty windbag who promised to put up this turkey, but in order not to be caught out in a cheap clownery has not gone beyond promises ... And in fact he looks like an adult ... SHADOW ...

 
dmitryigun писал(а) >>

You're falling for some bullshit ... It is thrown on the story, right above said ... The picture is nothing but "we have seen this" does not cause ... And call all sorts of crap can be kai-kai, or even non-normative - the essence of it does not change....IMHO - it's a fake, and Prival is just another empty windbag who promised to put out this turkey, but in order not to be caught in a cheap clownery has not gone beyond promises ... And a man judging by the photo like an adult ... SHADOW ...

Great. A very "profound" conclusion.

Make a better one and upload it to the public domain.

Some people already have versions of this indicator. And they can confirm that it doesn't redraw. But the work has not been completed. I have not made it as it should be. I have not yet mastered it.

I have already downloaded it and installed it on my computer and it is now ready for use. Got the sources via Skype without giving anything in return. Didn't deliver on his promise. And now you're puffing your cheeks because you can't figure out what's in there, how and why.

 
Prival >> :

Although you must be one of those who promised to work in a team with me. Got the sources via Skype without giving anything in return. Didn't deliver on his promise. And now you're puffing your cheeks because you can't understand what's there, how and why.

From now on you should be more selective in distributing your labours to pigs.)

 

Goertzel algorithm ( ssa close )

 
CROM писал(а) >>

Goertzel algorithm ( ssa close )

There's a good read on the algorithm here. It reminds me of our struggles)

http://am.ca/uve/qra_ve3ivm_4.html