1st and 2nd derivatives of the MACD - page 63

New comment
 
DmitriyN:
There are 365 days in a year, which means 365/7=52 weeks. There are 4 trading days (do not count Fridays). 4*52=208 days. Subtract the holidays. That will be about 200 days.


I can already see the logic here, or elements of it :)

Indeed, if you use the MA as a surrogate for the trend, it is logical to take the older CYCLE MA. The latter include daily, weekly, monthly and annual MAs.

Seasonal ones you can also use something like "day/night", but this is exclusive :)

 
To this day I don't understand - why use a surrogate trend instead of drawing a trend (a family of obvious straight lines)?
 
DmitriyN:

How can it be calculated? You can't. You don't know where the price will go during the next period. You cannot know over what period the pattern that your system is exploiting will be realised. However, you know exactly one thing - the regularity will be realized, because it simply can't not be realized.

In addition, you know how this pattern has been implemented before, you know the most difficult periods of its implementation and drawdown at these periods. You need to take this knowledge into account and at the same time take the necessary "safety margin".


So the pattern should be at the top of the list?

Where does one get it?

 
Zhunko:
I know! The author of the thread (Valera) needs to learn to understand food. This is also a speculum. By changing his food system, cleaning his organon, he will see the problem from a completely different angle. It is quite possible that the problem will disappear. The psyche and the somatics are inextricably linked.


I don't need you for a thousand years with your fuck-ups and detective games called find Valera, Vasya. You've got to be a human being. You're so advanced.
All your advancement is worthless if you have no humanity in our, mostly, shitty society,
in our already shitty society. (I don't mean forex or begging)

the rest of your problems have already been described in the smoking room, or rather not yours, but those of a very narrow group of people who have expanded their minds. and about the flip side of that. which no one talks about.

here it is https://forum.mql4.com/ru/47018/page78

 
Kocty:

2- How can you synthesise spectra from different things. At least how to get a synthetic spectrum from the bunch above.

Excuse me, can I ask you to formalise the problem?

Can you give us the definition of spectrum, its synthetics, pile and things?

in any notation you like...

;)

 

Can you tell me how to make a filter system like the one in the picture, by cutting them into parts. Do the green and red have to be phase reversed or what?

By shifting the filter backwards. In the pictures with the impulse response. https://forum.mql4.com/ru/45108/page41

https://forum.mql4.com/ru/45108/page44

 
These filters have such a thing as a decay period, i.e. you can set a decaying sine wave of impulse characteristic at the distance of 10 bars, and you can stretch this same part of the decaying sine wave to 50 bars, i.e. you can make this parameter dynamic. The volatility is recalculated ideally from the last bar (ideally from the last tick of course, but of course the compass is not strong enough).
 

There was a user's thread somewhere Balistica , so she was talking about interpolation, and someone tried to reproach her for actually talking about extrapolation, but calling it interpolation. The bottom picture shows not only TFs in the usual form, but also TFs increasing by 2 times and the intermediate values (for example - the most elementary values on the line between two points, and the number of these points is the number of minimal discrete steps of the smallest TF relative to these 2 points).

Looking at the picture you can see that the black lines are similar. Let's reject equal-volume (by tick volumes or by the number of points) bars.

It is clear that roughly speaking it and the same chart have different phases in different TFs. We may try to predict not the price itself, or rather the black line, but the difference between these black lines (it sounds crazy, but Makdi from one chart - Makdi from variations of the same period but from different TFs) and it follows that the forecast is not the price but the movement size. This phase line is different for each TF, we need to find a common one that's why macdi from acdi, etc. Д.

Or work on each phase for each TF separately.

So maybe for TF 1 the macdi has already crossed the zero line, but for TF 4 this same µdi (same period) is not yet. Thus, from these changes it is possible to predict the points of macdi intersection with zero line, and the information from one TF and from another one may be useful. What we do not have in one TF.

 
Kolysik:

In this way it is possible to predict the points of macdi intersection with the snapper from these changes and it may be useful to get information from one of the currents and from the other one. Which is not the case in one TF.


Of course it is possible. And what are we waiting for? Ai da piss grail!
 
Perhaps it is too early for the author to write a grail, here's more people informing at least so discussions may begin, who will improve the thoughts, who will write if he wants, then maybe in the base will appear, and then you can use it. BOA-GA-GA with a grain of salt. N or the comrade will himself deepen the thought. From the fact that there is no discussion seems to lose will be only those who are not profitable to discuss these things, judging by the reaction on the gestures of the author of such a place. By grail entry can be done from 2 sides - from the side of filtration as above (there certainly is more complicated, analysis of residues, and so on), but you can make a random process of this and approach from this side. The author will probably try to make the first option publicly available, for no one is willing to get in touch, and it will take him a long time to come to a final formalized variant of this approach. So maybe he decided not to waste time and immediately come to the second option, but he did not want to let such a volume of work go down the tube, so apparently the code base will be updated with interesting ideas, and since he has no formalization on mkl, he is not worried about it, let those who have it, especially as it turned out decent people among them are very few. Themes flashed many times, but did not go further than flickering, because for so many years in the code base there is not even a hint of something more or less decent in terms of decomposition. Let's wait and see what new has prepared for us the man with the same name, at least the end and a couple of posts with his attempts to crippling extrapolation I liked.
1...565758596061626364656667
New comment