Interesting topic for many: what's new in MetaTrader 4 and MQL4 - big changes on the way - page 72

New comment
 
Mischek:
Yeah ) by ZZ, if only to jack off.

The sum of the ZigZag-ideal segments for each individual day. )))

 
tol64:

The sum of the ZigZag-ideal segments for each individual day. )))

That's not what I mean . It's just a false goal, even as a "I won't catch up, I'll keep warm".
 
Mischek:
That's not what I mean. It's just a false goal, even as "if I don't catch up, I'll get warm".

It's still interesting. In my case, I consider any research initially just as practice in programming. And at the same time, along the way, my personal library of functions and indicators is replenished. The scheme can be useful later for visualization of some other ideas, so even in pursuit of false goals, you can sometimes find something useful.

That's why in this context - if I don't get it, at least I'll get warm. )))

 

Regarding 100 mbar/sec.

I paid attention to this

hrenfx: ...
  • The optimiser is designed to interrupt the run if the current state of the run does not meet the conditions (e.g. the drawdown is too high).
  • The MQL4 scriptprepares the history for the tester - it was written a long time ago

on this

Suppose you want to optimize your TS, and it is quite sensitive to the history data. To optimize it on pure tick history is madness (long). That's why it's worth the filter issue. For example, if you have a bad ping - you filter the ticks stronger, good - weaker. But besides such simple filter logic, you start to thin out ticks more and more, where on every section you see, how much the result of your TS differs from the reference one (without filter). Well, and you measure the run time after each thinning.

You can see that by thinning out e.g. 1000 times the ticks, you get a difference of 1% from the benchmark. At the same time, the testing speed increases correspondingly by 1000 times. Then you either stop at the choice of filter, or keep looking for the golden mean.

It is just a very competent tester will help you find such a golden mean automatically. And you can optimize your TS as quickly as possible (sometimes by several orders of magnitude), almost without affecting the quality of testing.

And here is this script

The script creates the initial symbol history file, on which the speed of testing/optimization of strategies on the"All ticks" model increases manifold, with identical results.
FastHistory - MQL4 Code Base
  • www.mql5.com
FastHistory - MQL4 Code Base: скрипты для MetaTrader 4
 
papaklass:

Yes, on a strong move, it's probably better to adjust the position at a not-so-best price (mark-to-market) than to fly away altogether. I just stupidly missed the moment of execution when I asked the question.

 
MT4 marketplace products will have to be mapped to mql5?
 
sovetnikmaker:
MT4 Marketplace products will have to be put on mql5?
A special site is already being prepared at MQL5.com - https://www.mql5.com/ru/market/mt4
MQL5 Маркет: MetaTrader 4
MQL5 Маркет: MetaTrader 4
  • www.mql5.com
Маркет - магазин программ для MetaTrader 5 и MetaTrader 4
 
And will all products be as tamper-proof?
 
Yes. There's already been a significant increase in protection, with the 509th build.
 

An interesting question has arisen here.

Will there be an extended date/time functionality in MT4 or not? I am asking about GMT in the first place.

1...6566676869707172737475
New comment