Author's dialogue. Alexander Smirnov. - page 21
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
to Mathemat
ow ow where are you?
Look at the previous page!!!
Korey, I don't really trust any code - either on Wasik or Easy - whose translation results in MQL4 are not confirmed by the author (who doesn't say anything about the code and can't until he posts a grad student here). We don't have a consensus on the correctness of translation to MQL4. And I'm very happy to finally see a number of figures that the author is unlikely to argue with since they're from his own article.
I've made a little effort and pasted figures from the table into the indicator (by typing them in directly) and this is what I've got:
Blue - SESS numbers from Excel table in Donetsk article, green - EMA(2, Close). Frankly speaking, I'm not interested at the moment whether it redraws or not. But how is this miracle fundamentally different from EMA(2)?
Below is the little indikator, with which I drew the blue curve.
2 Korey: I see, I see. I am trying to understand it.
P.S. Mr. Smirnoff compares the ideal signals with what the indicator produces, precisely based on what is drawn.
to Mathemat
Explanatory note))
When Mr. Smirnov was explaining the algorithm on his fingers, he bent his fingers in the wrong direction,
He said that he stroked both ways (the society was still discussing it).
So, when I saw the lack of right bars - well, I followed the explanation of Mr. Smirnov like a sheep.
But I looked again - not so but so and just expanded the array, so as not to plough through the entire algorithm.
So. HERE. My computer has 2 million cycles with double, but A-Revega m=400 counts almost 4 seconds.
That's like a Fourier sliding from a sample of 400.
I.e. this filter is a mathematical joke so far.
But the filter works as promised, the Remez method works, and D.V. Reevega is probably a strong mathematician too.
Here's an indicator that strokes both ways. But it is slightly overdrawn. But not by much. As an endpoint it works like a good muvinng. Rather it can be successfully used as a filter for Neural Network.
Unlike smart Caterpillar and Kalman, code is just a few lines. It works instantly. Compared to Wavlet, the stability, especially at the end, is several times better. Note that this is on a volatile pound.
because at increase (m) high-frequency component is oozing, so it does not smell of any Batterworth.
The Donetsk algorithm (A-Revega) is quite another matter, here, when increasing (m), not only the shiver goes away, but the wave is combed too,
and extremums remain in place (!).
You have a "poor trader's dream" in your picture, as unattainable as a GAZ-3110
to ANG3110.
You have "poor trader's dream" in your picture, it is as unattainable as GAZ-3110.
The Smirnoff1 & Smirnoff1p indices executed according to the Speculant do not correspond,
because at zoom (m) the high-frequency component is oozing, so it does not smell any Batterworth.
Another thing is the algorithm from Donetsk (A-Revega), here with increase (m) not only the trembling is gone, but also the wave is combed,
and extrema remain in place (!).
Yes, c.f. component in Speculator's variant doesn't go anywhere, and Donetsky algorithm is obviously smoother at large (computable with a gnash) m: at m=100 I have the terminal "hangs" almost all the time.
I apologize, Mr. Smirnov; I am too hung up on your explanations and on small m, on which, in fact, you have demonstrated everything.
The algorithm is obviously non-adaptive and strictly linear, with a constant window. Clearly it is not easy to calculate the filter k-ts, but to reconstruct them from the impulse function is quite a realistic task. Who of the DSP experts will undertake it? After all, this will optimize calculations by orders of magnitude!
And lastly, the comparison with Djurica is simply inappropriate here. Already with very small values of the parameter, Dzurik clearly wins in terms of speed, smoothness and underlapping. But at higher values, Djurik lags behind, though not by very much. These are different indicators.
And anyway, again on this A-Revega m=400 - found the EMA of 20 HLCC/4,
You're right, you can put a point on linear combinations.
If it's not too much trouble, so am I. Email in my profile.