a trading strategy based on Elliott Wave Theory - page 74

 
[quote]
For example, I threw in a script on history:
 
As you can see, after 4 bars we would have lost two channels, they just evaporated

 
Yep, given the more general view below, only 1 channel more or less matches. But I have one pass and I don't need to catch all of them - this is an indicator (expert) algorithm designed to work continuously on a rising history. Let's see what happens with my new method of channel selection, and then with "combat" (full-pass) application.
 
As you can see, after 4 bars we would have lost two channels, they just evaporated<br/ translate="no">
I'm afraid to be wrong, but they should, because if the script is running from another bar, the previous interval might not even be close to the new ones. I have the same, but the full capture of the shorter ones should (hopefully) have already happened before. With the longer ones we'll have to think and decide later.
By the way, I'm going to try using your experience of using objects for dynamic memory allocation instead of 2D arrays :)
 
Rosh:
And you have the reference point as some moving point in the past (firstBar), in relation to which the second channel bar (lastBar) slides in the future, am I correct?

Yes, this means a constant reference to the lastBar, which is what you need for an indicator or expert.
 
The picture is not much different on the clock

 
Rosh:
А у Вас оотправной точкой является некоторая скользящая точка в прошлом (firstBar), по отношению к которой скользит второй бар канала (lastBar) в будущем, я правильно понял? То есть , алгоритмы не совсем пересекаются и как бы разнонаправлены.

Yes, this means a constant reference to the last bar, which is what an indicator or expert needs.


I'm afraid we're misunderstanding each other again. In my picture firstBar is black, lastBar is green. It means that the binding is not to the last closed bar but to some bar in the past. Although maybe it's me who doesn't understand...
 
<br/ translate="no"> By the way, I'm going to try using your experience of using objects for dynamic memory allocation instead of 2D arrays :)


I only use two-dimensional arrays. I understand dynamic memory allocation in terms of Pascal, of course I don't have that. Just a parsimonious calculation algorithm.
 
By the way, I was wondering if we were violating Vladislav's recommended non-proliferation regime. Do you mind if I delete the details? If so, we can exchange them by post
 
I'm afraid we're misunderstanding each other again. In my picture firstBar is black, lastBar is green. So you are not linking to the last closed bar, but to some bar in the past. Although maybe it's me who doesn't understand... <br / translate="no">

The direction doesn't matter, the important thing is that the outermost lastBar (or firstBar) is always the same distance from the last bar. And I get the first and the last one mixed up all the time myself :)
Reason: