From theory to practice - page 447
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
I think the correct sliding window = 24 hours. This just needs to be confirmed experimentally.
It's like working on the M5 with a swing period of 288, which apparently some bas is doing, chopping up dough on the Bollinger Bands. That's all I could make out from his sheets that he's rolling around in here.
Well, the logic is there, but I can say that there is one pattern in the market, which goes from day to day - it's an intersessionary flat-out.... are you sure you should analyze this pattern?
As I'm not much of a theorist and not much of a practitioner, but I happen to write EAs upon my request and for a fee. Well, the customers who really want to trade with EAs all ask me to allow them to trade at certain times. That is basically all I am thinking about.
Distribution of real tick quotes in sliding time window = 4 hours:
Histogram presented for 500,000 values.
No Poisson flow...
Therefore, starting next week I am moving to window = 8 hours.
Will try to make you a tick collector. Tired of reading how you are agonising with this.
I'm about to learn something....:
https://www.mql5.com/ru/forum/265056
Conclusion: to move to a nearly stationary process and variance = const, a move to a sliding window of higher dimensionality is necessary.
Everything.
If a larger window, it is possible to move to minutes.
Conclusion: to move to a nearly stationary process and variance = const, a move to a sliding window of higher dimensionality is necessary.
That's it.
Again another obscurantism - this stationary process with reports "once a year, when the dog whistles" has nothing to do with real input BP.
That's why the price is zero, because we still need to trade according to the original process.
In short, we have "Woe from Wit" in all its glory, with the main actor absurdly dancing the "Swan Lake" dance. ))
Again another obscurantism - this stationary process with the "once a year, when the cows come home" reports has nothing to do with real input BP.
That's why the price is zero, because you still have to trade according to the original process.
In short we have "Woe from Wit" in all its glory with the main actor absurdly dancing the Swan Lake dance. ))
The mind is glasses, which the monkey doesn't know where to put it). But nevertheless, by trial and error, something may work out. There are theoretical physicists and experimental physicists. Alexander is more of an experimenter.
Channel and increments with a period of 60 minutes.
And here is how the price was
As you can see the price came back, though before that it had gone down about 117 five-digit points.
And this is what the density chart looks like at the time of the breakout
Here's something else you need, like velocity or inertia, maybe something like this
de Broglie waves.
https://ru.wikipedia.org/wiki/%D0%92%D0%BE%D0%BB%D0%BD%D1%8B_%D0%B4%D0%B5_%D0%91%D1%80%D0%BE%D0%B9%D0%BB%D1%8F
Only for me it's beyond comprehension.
I wanted to calculate the angles of a triangle, started calculating the hypotenuse (why I thought the triangle was right-angled, but who knows) and came up with the following.
The data:
d - value of variance at a given moment
tn - period time in minutes transformed in some way
Hypotenuse^2 = d^2 + tn^2
Conclusion: Sqrt(Hypotenuse) = d
We obtain a triangle with two equal sides equal to d and one side tn, it is clear that there is no 90 degree angle.
From this data you can calculate the angle between the sides of d and get some dispersion angle, and between the sides of d and tn. The angle between the sides of d is much smaller.Now watch the wonders.
Let's see what the probability distribution of the sum of tick increments in the sliding window = 8 hours looks like (on the same data on which I built the histogram for the sliding window = 4 hours):
Its statistics:
We see that this distribution is already practically a Gaussian distribution.
There is an opinion that already in the sliding window = 12 hours, we will observe a stationary Ornstein-Uhlenbeck process with a "return to the mean".
Hallelujah brothers!!!
GRAALNASH!
I feel with my spine that the distribution should be the same at any sliding window. Why I don't know. Maybe it's like looking at an oil painting from different distances, from afar it's clearer than from up close, but the essence doesn't change.