From theory to practice - page 936

 
Martin Cheguevara:

You're right... but maths is maths in Africa.... in 1900 too... what's the point of a genius if he can't perform a miracle))

=D

they were solving applied problems. They did perform a miracle. Only you didn't notice.)

 
Maxim Kuznetsov:

the simplest implementation of this method you have done yourself many times :-) at the very beginning...

the closing loop for(int pos=0;pos<total;pos++) instead of the modern way for(int pos=OrdersTotal()-1;pos>=0;pos--)

then the orders are closed through one which is close to the desirable one and the drawdown is less but the procedure has to be repeated until there is nothing left at all

and this is from the old to the new ones. From new to the old ones, it will be for(int pos=OrdersTotal()-1;pos>=0;pos-=2) and should be repeated until all of them are closed.

PS/ that is, when the need and opportunity to "unload" the grid arises, the orders are closed in 1. The direction of passage is chosen according to taste or the position of the stars

In addition, these functions must be checked and checked again by prices, ticks and time. Nothing should be left to chance in MQL, not even a declaration of the double type.

Igor Makanu opened my eyes to the fact that double n = 0; may suddenly become int or even become anything XD.

 
Yuriy Asaulenko:

Applied problems were solved. They did perform a miracle. Only you didn't notice)).

Well, of course I am not arguing about applied problems))

But correct and objective - that is uniquely correct recognition of price movements in the markets - is far from being an applied problem... unfortunately...

 
Yuriy Asaulenko:

Applied problems were solved. They did perform a miracle. Only you didn't notice).

By the way, I already built the methods of the respected Kolmogorov for artillery guns into my programs. And it worked well, except for one thing - the load on the deposit was frantic, especially from the commission...

And when I understood his methods, he used area methods of probability calculation, and it was similar or even very similar to Bayesian methods... I may not have read somewhere that he was paying attention to his methods...

 
Martin Cheguevara:

Of course, I am not arguing about applied tasks...)

But correct and objective, i.e. unambiguously correct recognition of price movements on markets, is far from being an applied problem... unfortunately...

The problem is applied, but it is impossible in principle. Even Kolmogorov's 41 conditions do not correspond to market quotes. And what was the point of spinning it? And what, in fact, was there to be disappointed about? The article is not about us at all.)

 
Yuriy Asaulenko:

The problem is applied, but it is impossible in principle. Even the conditions of Kolmogorov 41 market quotes do not correspond in any way. What was there to spin? And what, in fact, was there to be disappointed about? The article is not about us at all.)

If you analyse not a continuous function, but some square object with price movements in it as a target and orders as projectiles.)) You can attach anything to the market.)

This is what I was doing before and thanks to it I thoroughly studied many works dedicated to mathematics. At least it was of some use).

 
Martin Cheguevara:

If we analyze not a continuous function, but an area object with price movements in it as a target and orders as projectiles.))

If you have read the article till the end, the condition there is not stationarity at all but a limited frequency spectrum of the process. An aeroplane corresponds to this condition but market VR has an unlimited spectrum. That is all. These methods do not work.

Some other methods are possible, I do not know.

 
Yuriy Asaulenko:

If you read the article to the end, the condition there is not stationarity at all, but a limited frequency spectrum of the process. The aircraft meets this condition, but the market VR has an unlimited spectrum. That is all. These methods do not work.

Some other methods, perhaps, I am not aware of.

I understand that. In general, it is clear that pulling everything out of the market charts is wrong.)
An aeroplane cannot suddenly be 25km away from where it was a moment ago.)
 
Martin Cheguevara:

Which way does the grid work for you?

on a bounce or following the trend?

my grid works "on the button" :-)

When I decide it is time to buy, I push the button and the robot sets the grid. That is, when I see a reversal in several pairs at once.

 
Maxim Kuznetsov:

My grid works "on the button" :-)

When I decide it's time to buy - I press the button, the robot sets the grid... Mostly I enter a "counter-trend", based on a multisymbol indicator. That is, when I see a reversal in several pairs at once.

It's true) Knopul:B is funny))) It turns out that you enter based on a posteriori information. It's good)