From theory to practice - page 141

 
bas:
dukascopy.com or ticks.alpari.org
I will try this weekend.
 
Alexander_K2:

Man, I'm tired of repeating that there are NO MACHINES!!! Open a trade when the average value of increments (bottom graph) crosses the 99.5% confidence level, and close when this average value of increments becomes = 0.

Compare the lower graph of the average value of increments with the upper graph of the price itself.

And you subtract the bottom graph from the top graph, and apply the result to the price graph. This will be the implicitly implied swing that supposedly does not exist.

 
bas:

And you subtract the lower chart from the upper chart and apply the result to the price chart. This will be the implicitly implied swing that supposedly does not exist.

????????!!!!!!!!!!! I'll try..........
 
Alexander_K2????????!!!!!!!!!!!
What's so surprising here? You decompose the graph into the sum of the trend component and the cyclical component.
 
By the way, the average increment is simply (Ask[i] - Ask[i+N])/N, you can make the calculation easier.
 
Alexander_K2:

Man, I'm tired of repeating that there are NO MACHINES!!! Open a trade when the average value of increments (bottom graph) crosses the 99.5% confidence level, and close when this average value of increments becomes = 0.

Compare the lower graph of the average value of increments to the upper graph of the price itself.

A VERY robust robust system. No need to endlessly search for the right average. I do not see loss-making trades in 4 days this week on AUDCAD pair from the word "at all".


And I thought that this is an average)

What is a mashka? Is it the one you use for waxing the floors in the barracks or the one you use for boning?

 
Sergey Chalyshev:

I thought MASHKA was the average.)

What is a mashka? Is it the one they use to wipe the floors in the barracks or the one by the thigh?

Well, it was you who introduced the average for the increments and I fell for it.

In this case, it is still the sum of all values of increments (both positive and negative) for a certain sample size. That's more accurate.

Man, I wish I hadn't replied to comments at all - now people get confused with the text because of people like you... He's got double vision, or he'll broadcast some other shame. Ugh...
 
Alexander_K2:

Well, it was you who introduced the average value for the increments, and I fell for it.

In this case, it is still the sum of all values of increments (both positive and negative) for a certain sample size. It is more accurate.


It is amoving average (moving average), in common parlance from the abbreviation MA.

It's just a series to which the algorithm of MA processing has been applied.

Take the cumulant of MA taken from the series of increments and you get MA from the cumulant.

 
Alexander_K2:

Well, it was you who introduced the average value for the increments, and I fell for it.

In this case, it is still the sum of all values of increments (both positive and negative) for a certain sample size. This is more accurate.

More accurately, it is a scale with a certain period (standard indicators have this).

But positive and negative values cannot be used in a scale, especially modulo.

 
Alexander_K2:
I, alas, do not have archived data of this depth....
On 11 January I collected this data for you, https://www.mql5.com/ru/forum/221552/page128#comment_6327590.