a return to the fairy tale - page 4

New comment
 

The situation seems to be this way: on the dukes there are 4 digits ( and fractions ) and on the nds there are 5...so in the code place-

if(currentSymbolOrderPos < 0)
{
if(priceUp < ask-p)
{
up = up + 1;
priceUp = ask;
if(TimBoolUp == false)
{
TimeSpeedUp = TimeCurrent();
TimBoolUp = true;

}

there is a natural filtering with threshold p ( hysteresis )...

it is possible that adding a similar value of such threshold to the μl code would remedy the situation in mt...

Files:
woc.experiment_porog.mq4  8 kb
 
Please tell me the settings(EA input parameters, symbol, TF, spread, broker) at which the FOC in the tester shows graality.
 

default (the exp. above)...anyf ( it's on ticks )...eur -usd... alpari andd

February:


Chart mismatch errors 0
Initial deposit 1000.00
Net profit 40153.45
Total profit 50140.24
Total loss -9986.79
Profitability 5.02
Win expectation 102.43
Absolute drawdown 8.64
Maximum drawdown 2223.76 (5.28%)
Relative drawdown 5.28% (2223.76)
Total transactions 392
Profitable transactions (% of all) 293 (74.74%)
Loss transactions (% of all) 99 (25.26%)



 
atik:
Well, in this one I have inserted the opening of the bar... &&TimeCurrent()==Time[0]... so there is only one value left biased in the tester ( mt synthesis ) it is the value at the moment Speed

Such a condition is also triggered when the bar is not open. Example:

void start()
{
  static int PrevTime = 0;
  static int i = 0;
  
  int P = Period() * 60;
  
  if (Time[0] != PrevTime)
  {
    PrevTime = Time[0];
    i = 0;
    
    Print("NewBar: " + TimeToStr(PrevTime, TIME_SECONDS));
  }
    
  i++;
  
  Print("Tick " + i + ": " + TimeToStr(TimeCurrent() % P, TIME_SECONDS) + ", " + DoubleToStr(Bid, Digits));
      
  return;
}

P.S. Replace your condition with this one:

... && Volume[0] == 1)
and see the results.
 
hrenfx:

Such a condition is also triggered when the bar is not open. Example:

P.S. Replace your condition with this:

and you'll see the result.

Why would you want to insert a volume equal to ==1 ??? you could remove this condition altogether, by the way...

it makes sense to introduce a threshold on volumes : 

... && Volume[0] > PV )
 

This condition is only fulfilled at the bar opening.

P.S. The grail in the tester turns out to be exactly that.

 
hrenfx:

This condition is only fulfilled at the bar opening.

P.S. The grail in the tester turns out to be exactly that.


So there is a logical solution: why not limit the entry time by the time of bar opening (or e.g. after the bar opening time at the 1st or 2nd or 3rd tick) and synthesize the previous ticks in the same way as with mt (and take these synthetic ticks for comparison instead of real ones)?
 
The grail in this case would have worked if the opening of positions had taken place at the beginning of the bar.
 
hrenfx:
The grail in this case would have worked if the opening of positions had taken place at the beginning of the bar.

i.e. do you think it is possible to create a real system similar to the tester system in principle using the above-mentioned principle of its creation?
 

You have made a perfectly valid attempt to get rid of looking ahead in this EA - to open only at the beginning of a bar. If such an opening were correctly implemented, this EA would show profit, then it would be just as suitable for real. All one would have to do is to simulate the synthesis of tester ticks.

All this, of course, is nonsense. It is better to check such ideas on tick data from the beginning.

1234567891011
New comment