Новая версия платформы MetaTrader 5 build 1930: Плавающие окна графиков и .Net библиотеки в MQL5 - страница 86

[Carl Schreiber]

Slava:
Just do it

Thanks!

[Alain Verleyen]

Vladimir Pastushak :

Подтверждаю, противный и неудобный баг...

Это старый. И есть несколько таких же.

[Carl Schreiber]

The StratgyTester puts me into troubles:

I am using OnTester() for the optimization (gen. algo., custom max) to give more weight to the Profit then the normal way of Profit/DD%:

double OnTester() {
   double dd = sqrt(TesterStatistics(STAT_BALANCE_DDREL_PERCENT)),
          pr = TesterStatistics(STAT_PROFIT);
   if (dd == 0.0) return(100.0*pr);
   return ( pr/dd );
}

Here are the relevant columns of the first two lines of the table of the results of the optimization:

Result	Profit	DrawDown
5.892,29	19.421,04 €	11,78 %
5.834,81	21.086,40 €	9,42 %

How can the result of the first line be bigger with a smaller Profit and a higher DD% the the result of the second line?

This is what I get when I calculate the result myself in Excel:

Result	Profit	DrawDown	SqrRoot(DD)	Pr/SqrR(DD)
5.892,29	19.421,04 €	11,78 %	34,32 %	56.584,81 €
5.834,81	21.086,40 €	9,42 %	30,69 %	68.703,20 €

Did I miss something?

Note: the value of the result is not my problem, but I don't understand that the result of the first line is bigger than the one of the second line, so the relation of the results seem to be false!

[Rashid Umarov]

Carl Schreiber:

Result	Profit	DrawDown	SqrRoot(DD)	Pr/SqrR(DD)
5.892,29	19.421,04 €	11,78 %	34,32 %	56.584,81 €
5.834,81	21.086,40 €	9,42 %	30,69 %	68.703,20 €

Did I miss something?

Note: the value of the result is not my problem, but I don't understand that the result of the first line is bigger than the one of the second line, so the relation of the results seem to be false!

Sqrt(11) = 1.8 3.3

Sqrt(0.11) = 0.33

It is answer

Some mistake was happened to my calculator

[Carl Schreiber]

Hmm? - so:

TesterStatistics(STAT_BALANCE_DDREL_PERCENT);

returns values for 'all' and 'nothing' between 0.0 and 100.0 instead of 0.0 and 1.0?

In this case it should be mentioned in the reference as normally nobody would expect this - I guess.

But even in this case I get this table:

Result	Profit	DrawDown	SqrRoot(DD)	Pr/SqrR(DD)	Mt5-DD	Sqrt(Mt5-DD)	Pr/(Sqrt(Mt5-DD)
5.892,29	19.421,04 €	11,78 %	34,32 %	56.584,81 €	11,78	3,43	5.658,48 €
5.834,81	21.086,40 €	9,42 %	30,69 %	68.703,20 €	9,42	3,07	6.870,32 €

Even in this case line 2 would be better than line 1 - in contrary to the result of the StratgyTester.

1. I (b1961) use sqrt for the DD% only once as sqrt(11) = 3.3, but sqrt(sqrt(11)) = 1.8211.
   
2. The €-Sign for the profit is 'introduced' by Excel while I optimize EURUSD on an USD-Account ( in case of doubts).
   

[Rashid Umarov]

Carl Schreiber:

Hmm? - so:

returns values for 'all' and 'nothing' between 0.0 and 100.0 instead of 0.0 and 1.0?

No , I was wrong. Fixed my previuos post, sorry

[Carl Schreiber]

Rashid Umarov:
No , I was wrong. Fixed my previuos post, sorry

Ok - forgiven ;)

But what is with my problem?

[Rashid Umarov]

Карл Шрайбер :

Ок - простил;)

No guess. Try to use this article in order to find a reason - https://www.mql5.com/de/articles/746

[Rashid Umarov]

Carl Schreiber:

Ok - forgiven ;)

But what is with my problem?

No guess. Try to use this article in order to find a reason - https://www.mql5.com/de/articles/746

[Carl Schreiber]

Rashid Umarov:
No guess. Try to use this article in order to find a reason - https://www.mql5.com/de/articles/746

I'll try - is there a way to get the actual number of the pass and/or the generation in order not to print> 10'000 lines?