Random Flow Theory and FOREX - page 16
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
The sum is an integral only for a discrete value. In order to decompose a Fourier series function, we take the last n samples. This is information from the past. Here is more details about the energy and power of the signal.
P.S. The epigraph here is brilliant ;-). If you have read it and then you think that simple military radio engineer has answers to all questions ;-), you are mistaken.
Great, I love it:
Математик может говорить все, что взбредет ему в голову, но физик обязан сохранять хотя бы крупицу здравого смысла.
Josiah Gibbs. American physicist, 19th century.
I would like to know where that very grain of common sense in the energy of an information signal is, if information itself is considered an intangible concept.
Margarita Piastolova. Irkutsk geophysicist of Ural school, the twentieth century.
What we are trying to do here is exactly an attempt to apply physical notions to a purely informational object. I myself have a theoretical attempt to describe rate motion (10 years ago; who would have thought then that 10 years later I will do the same rubbish?), and there energy arises too (exactly energy, rather Lagrange F), but on an entirely different basis: no spectral analysis can be smelled there.
Prival, the sum of complex spectral components is nothing but the Fourier transform (PF) itself from our discrete series. And the sum of the squares of their amplitudes is what is commonly called signal energy. So, what are we counting at the output of the indikator - the signal energy or its PF?
I must be an idiot for asking such questions, Mr. military radio-engineer?
P.S. In this study I am trying to act as a physicist.
I will try to retain at least a modicum of common sense.
The discrete Fourier transform is a transform that decomposes the original function into basis functions, which are sinusoidal functions, that is, it represents the original function as a sum of sinusoids of different frequency, amplitude and phase. On the complex plane, it is a pile of vectors.
By definition, the sum of two complex numbers z1=a + ib and z2=c + id is the complex number z1+z2=a + c + i(b + d). Now suppose that the red vector in the figure is the sum of all Fourier transform vectors. The indicator outputs the projection of this vector on the y-axis (Re).
I.e. it is not PF (but its part) and not energy, but some quantity rigidly connected with these concepts. Maybe more accurately it is a force that pushes the curve up and down. After all, energy does not just exist, it is always directed somewhere.
I will allow myself to participate as a physicist.
About the epigraph and all other thoughts about energy and information.
Any concept is an immaterial thing, just because it is a concept, not a phenomenon itself. Information is a non-material phenomenon. Perhaps this is what the Irkutsk geophysicist had in mind. If so, she is undoubtedly right. It is not clear only why she does not see sense in "energy of an information signal". I, truth, do not know what "an information signal" is and how it differs from a simple signal, but probably the woman meant a signal which bears the necessary information. Why cannot this signal have energy, if any signal has energy, regardless of whether it carries information for us or not? Energy is a universal characteristic of physical phenomena and all phenomena of material world possess energy. It is our merit that we have adapted some of these phenomena for transferring information. So there is no need to confuse a carrier and what it transfers. I think that a geophysicist has no problem with her notebook, where she keeps non-material information about telephone numbers.
Likewise in forex. Price is, of course, information. But it reflects real physical processes with the flow of money - quite tangible energy that people use. I can even tell you the type of that energy - potential energy. And there is no conflict here, just as there is no conflict between the immaterial speedometer reading and the physical speed of a material car.
Prival' s idea to draw a parallel between the movement of the price and the movement of an aeroplane may not be so obvious, but it does not contradict any physical ideas. Imagine the movement of an aeroplane (neither too heavy, nor too fast) in a turbulent flow of air. I think it is quite similar. And the concepts of energy, inertia, etc. are quite applicable here, although it must be done carefully and competently.
By definition, the sum of two complex numbers z1=a + ib and z2=c + id is the complex number z1+z2=a + c + i(b + d). Now suppose that the red vector in the figure is the sum of all Fourier transform vectors. The indicator outputs the projection of this vector on the y-axis (Re).
I.e. it is not PF (but its part) and not energy, but some quantity rigidly connected with these concepts. Maybe more accurately it is a force that pushes the curve up and down. After all, energy doesn't just exist, it is always directed somewhere.
Energy is a scalar. It exists for a reason, but it is not directed anywhere, unlike force. The presence of force tells us about the flow of energy from one form into another, while its direction is only about the geometry of interaction between objects.
As far as I know, the projection onto the real axis isolates the cosine-transform related part from the FP. So maybe from the beginning not to do all FP, but only a part of it ?
Gentlemen, let me give you some advice too...
P.S. I lost half a day again, saying the obvious to me. I have not yet reached my goal :-(. I have not found those who really want to help me in the creation of such TS. I'll probably delete the hell this thread.Well, I think I offended you again by inadvertence. I apologize if it's wrong.
IMHO: we all "really" help each other here. The forum has taught me a lot. If I've made my points frankly (with no thought of criticism), it's up to you to take them into account or discard them. I understand your impatience to start "shaking things up", but more time can be lost later if "on shore" you don't think it over well and even leisurely.
The AI references you cited are examples of not the best attempts at artificial neurons in TC. Now, if you look specifically at books on AI or NN, you'll see that what's given in those links is primitive. By the way, advanced Kalman filters are also implemented on NN, for example.
I don't think we are ready to start writing something in MQL4, although I assure you it won't be a big problem to build the Kalman filter. The main task, which you have outlined - the prediction of the price trajectory - depends, as you have rightly said, on the task of pattern recognition (patterns, trajectories, groups of signals, indicators - whatever). I have no doubt that you are sufficiently well versed in this problem. But it is generally accepted (and Matemat wrote about it here) that NN is very successful in solving such problems, much more efficiently than many other tools.
OK, again, I seem to be "loading". Prival, as the owner of this thread (at least, not to destroy it because of me) - you may hint about undesirability of my presence here, and I promise: I'll be silent like a fish :-).
Not sure it's what I need yet, and not sure it's right, but it worked!
Let Prival post the code himself (if he wants), and I'll brag about the picture ;)
Mathemat Yurixx Neutron
You mathematicians are cooler than me. It may be so. In the indicator, if you enter h=false, the projection on the other axis is plotted (if I understand correctly it is the sine PF lower chart). But that's how to interpret it I don't know, something to do with phase (but it's like I said before a thing in itself). If I am right that it is a force (F), then it would be interesting to see how the mass changes. F=m*a. Acceleration seems to be there too. can you build an indicator ?
rsi
Don't go away. Very glad my "nonsense" is starting to appeal. And perhaps there is some common sense in my statements. If you don't mind, program a Kalman filter in MQL. I have laid out the formulas above. I can give you a more detailed (step-by-step) instruction. You are a master and you can program it in 1 hour but I will need 24 hours and I will not be able to program it in MQL for three days.
komposter
I think I made an error when calculating y(x)=a*x+b. I will attach the image and assure you that it is calculated correctly and I will certainly post the indicator.
I am online now.
Edit, when you choose h=false it is not the projection, but the sum of the angles it seems.