create an expert for mt4 using a programme made in exel - page 26

New comment
 
alsu:
In the foreground, short-term significant harmonics are a dime a dozen, just look at the graph.
The problem: they don't tend to keep their periods in a chaste state. I.e. periodicity is there, but too @fucking much. ;)
 
alsu:

25 again.

You gave us an example with separation and validation of significant harmonics - what prevents us from using them for short-term forecasting of non-periodic processes? We don't consider a signal to be a periodic function, nor do we consider its spectrum stationary at all, but we do imply that it contains certain harmonics with an amplitude varying slowly enough to solve the prediction problem for several counts ahead. Or do you think that Fourier will not work here as well?

Yes, you can apply this method however you like.

In any case, any function approximation uses the inertia property of the process described by the function to extrapolate - that is, to predict function values beyond the right-hand edge. Which I think is the flaw in Fourier: it converges to a priori known values of the left-hand edge in the limit (given an infinite number of harmonics). Which isn't exactly what we're aiming for when trying to predict the market.

Good luck.

 
MetaDriver:

To tell you the truth, I don't think it will. For a long time I think so, but I still could not formulate my feelings. Vladislav summed up my vague thoughts very clearly. Right in the hole.

// 2 VladislavVG by the way, thank you!


I love you know mine and I "feel" it.

Take and find, for example, in the foreseeable future (for a short-term forecast :) equality of sections of the "total" function - starting periods of 11 23 and 51 bars respectively...

;)

By the way, when choosing significant frequencies, all other things being equal, preference should be given to those which have period - prime number. Practice of this simple rule often allows to avoid drift of frequencies.

 
MetaDriver:
Problem: they don't tend to keep their periods chaste. I.e. the periodicity is there, but it's too @fucking much. ;)
And who said it would be easy? But the thing is that there are simpler processes than Forex, and it works there in spite of the sensations. At Forex it also works, but not always, because of the @fuckness. This is why our task of forecasting is complemented by the task of detecting the situations when there is no need to make forecasts. We are not soldering a receiver to demand a whole signal at the output, pieces are enough for us, at least sometimes))
 
Sorento:

By the way, when selecting, all other things being equal, essential frequencies, preference should be given to those with a period - a prime number. The practice of applying this simple rule often avoids frequency drift.

didn't know, although there is a rationality in it - we can cut off some of the non-linear interference....
 
alsu:

a little on the topic of the branch)))

It's about "Right in the hole" I was thinking...

Sort of an "oszuzzeniya" - don't fail me.

;)

 
Sorento:


You asked for help in writing an EA. I asked - "program" in Excel in VBA?

There was no response. But there were tablature and heated discussions.

Is there a need for translation?

And for Excel (even if there's no VBA and just a chain of formulas) it would be understandable without explanations.

I doubt that an article (if there is one) will clear up all the questions...

;)


Sorry, I'm not familiar with VBA at all, but calculations, conditional and unconditional transitions, definitions of robot behaviour strategy, possible price change limits in the foreseeable future, extraction and subsequent analysis of sufficient information to ensure break-even trading are made by the robot through a chain of formulas located on 25 lines of excel. The only source of information for the robot is line 26 - input of the initial data, and the robot does not care about which currency pair to work with and when to start working - it is determined by the owner.
 
Vinin:


The article will not, of course, remove the questions, but will only create new ones.


the article will give you the inner confidence to understand the theoretical assumptions behind the robot's actions, of course questions will arise, but apparently they will be constructive and useful for both sides
 
Vinin:


I am not touching the theory itself (I have the report, not the article).

I just want to try to make an indicator. And the formulas themselves are not enough for that.


And you will find out the know-how as a result of discussion, you can't squeeze everything into one report. I suggest that we abandon the qualification of what we create together as an indicator, and stop at the concept of "robot".
 
yosuf:

the article will give you inner certainty in understanding the thertical presuppositions behind the robot's actions, of course there will be questions, but apparently they will be constructive and useful for both parties


Inner confidence?

Oh, cool!

Can it be deciphered?

Or is it like - once it's published, it's right?

1...1920212223242526272829303132
New comment
Reason: