Errors, bugs, questions - page 140

[Test Account]

jmp:
I set the size and location of the charts, close MT5, open it - as a result all the charts are stretched across the width of the window, how do I get rid of it?

Create a request to servicedesk.

Please provide screenshots before restarting the terminal (with location settings) and after restarting it.

[Test Account]

Cherrr:

From this rambling story I have only one question: is there any way to deal with this?

You need a detailed description of the testing parameters and dates with and by which you set but don't get the desired result. It would be very helpful to have an expert that you run.

This is best done in the form of a request to servicedesk.

[[Deleted]]

Developers.

There are no symbols (or just not visible) in the market overview of the new build. Or is it just me who has this glitch?

Bild is running under WinXP 32 bit.

[Yedelkin]

Interesting:

Or is it just me?

The build is running under WinXP 32 bit.

I still have it all. XP. Right mouse button doesn't help?

[[Deleted]]

Yedelkin:
I still have everything. XP. Right mouse button doesn't help?

After three restarts, everything seems to have appeared. I'll try it on other terminals...

PS

It's OK on other terminals. Probably a random glitch :(

[Aleksey]

Curious. Has anyone noticed a trick like this?

After optimisation, we select the option (best) on"Run Single Test" to check and view the graph,

and in the single test, we get completely different results

And in the settings everything is the same as in optimization.

I've already caught it twice, I stopped optimization, I start it, and the test doesn't match the optimization, and from +9000 to minus 9000... I don't understand what the problem is. Maybe it was mine. I haven't contacted the service yet.

I still have more to add. In the last case:

After optimization (before testing), changed test method from'OHLC to M1' (was during optimization) to'Every tick', the result was the same negative. Then changed the test method back to'OHLC to M1' and tested again. The differences are insignificant, those parameters on which there was optimisation are given.

[Sergey Gritsay]

I have written a function to check the trading session. On the demo works in the tester does not want to, printed the data in the log is what I got

From the demo:

NN      0       Проверка (EURUSD,M5)    00:49:35        2010.09.21 22:49:32 >= 2010.09.21 00:00:00 && 2010.09.21 22:49:32 < 2010.09.21 23:59:00
DI      0       Проверка (EURUSD,M5)    00:49:37        2010.09.21 22:49:33 >= 2010.09.21 00:00:00 && 2010.09.21 22:49:33 < 2010.09.21 23:59:00
LI      0       Проверка (EURUSD,M5)    00:49:37        2010.09.21 22:49:34 >= 2010.09.21 00:00:00 && 2010.09.21 22:49:34 < 2010.09.21 23:59:00
QH      0       Проверка (EURUSD,M5)    00:49:38        2010.09.21 22:49:34 >= 2010.09.21 00:00:00 && 2010.09.21 22:49:34 < 2010.09.21 23:59:00
KH      0       Проверка (EURUSD,M5)    00:49:41        2010.09.21 22:49:37 >= 2010.09.21 00:00:00 && 2010.09.21 22:49:37 < 2010.09.21 23:59:00
CK      0       Проверка (EURUSD,M5)    00:49:41        2010.09.21 22:49:38 >= 2010.09.21 00:00:00 && 2010.09.21 22:49:38 < 2010.09.21 23:59:00

From the tester:

RE      0       Core 1  00:40:08        2010.01.04 04:01:00   Условие 2010.01.04 04:01:00 >= 2010.01.04 00:00:00 && 2010.01.04 04:01:00 < 2010.01.04 00:00:00 не выполнено
RP      0       Core 1  00:40:08        2010.01.04 04:01:02   Условие 2010.01.04 04:01:02 >= 2010.01.04 00:00:00 && 2010.01.04 04:01:02 < 2010.01.04 00:00:00 не выполнено
JO      0       Core 1  00:40:08        2010.01.04 04:01:04   Условие 2010.01.04 04:01:04 >= 2010.01.04 00:00:00 && 2010.01.04 04:01:04 < 2010.01.04 00:00:00 не выполнено
JJ      0       Core 1  00:40:08        2010.01.04 04:01:06   Условие 2010.01.04 04:01:06 >= 2010.01.04 00:00:00 && 2010.01.04 04:01:06 < 2010.01.04 00:00:00 не выполнено
RI      0       Core 1  00:40:08        2010.01.04 04:01:08   Условие 2010.01.04 04:01:08 >= 2010.01.04 00:00:00 && 2010.01.04 04:01:08 < 2010.01.04 00:00:00 не выполнено
GD      0       Core 1  00:40:08        2010.01.04 04:01:10   Условие 2010.01.04 04:01:10 >= 2010.01.04 00:00:00 && 2010.01.04 04:01:10 < 2010.01.04 00:00:00 не выполнено
KP      0       Core 1  00:40:08        2010.01.04 04:01:12   Условие 2010.01.04 04:01:12 >= 2010.01.04 00:00:00 && 2010.01.04 04:01:12 < 2010.01.04 00:00:00 не выполнено

Code:

//+------------------------------------------------------------------+
//|                                                     Проверка.mq5 |
//|                        Copyright 2010, MetaQuotes Software Corp. |
//|                                              http://www.mql5.com |
//+------------------------------------------------------------------+
#property copyright "Copyright 2010, MetaQuotes Software Corp."
#property link      "http://www.mql5.com"
#property version   "1.00"
MqlDateTime     str;
//+------------------------------------------------------------------+
//| Expert initialization function                                   |
//+------------------------------------------------------------------+
int OnInit()
  {
//---

//---
   return(0);
  }
//+------------------------------------------------------------------+
//| Expert deinitialization function                                 |
//+------------------------------------------------------------------+
void OnDeinit(const int reason)
  {
//---

  }
//+------------------------------------------------------------------+
//| Expert tick function                                             |
//+------------------------------------------------------------------+
void OnTick()
  {
//---
   sesion(_Symbol);
  }
//+------------------------------------------------------------------+
bool sesion(string sym)
  {
   datetime start,finish;
   string d1,d2;
   datetime date=TimeCurrent();
   TimeToStruct(date,str);

   string ty=string(str.year);
   string tm=string(str.mon);
   string td1=string(str.day);

   SymbolInfoSessionTrade(sym,(ENUM_DAY_OF_WEEK)str.day_of_week,0,start,finish);
   d1 = TimeToString(start,TIME_MINUTES);
   d2 = TimeToString(finish-1,TIME_MINUTES);

   datetime dd1  = StringToTime(ty+"."+tm+"."+td1+" "+d1);
   datetime dd2  = StringToTime(ty+"."+tm+"."+td1+" "+d2);

   if(date>=dd1 && date<dd2)
     {
      Print(date," >= ",dd1," && ",date," < ",dd2);
      return(true);
     }
   Print("Условие ",date," >= ",dd1," && ",date," < ",dd2," не выполнено");
   return(false);
  }
//+------------------------------------------------------------------+

[Vladimir]

pronych:

Curious. Has anyone noticed a trick like this?

After optimisation, select the (best) option on"Run Single Test" to test and view the graph,

and in the single test, we get completely different results

And in the settings everything is the same as in optimization.

Here's the two times I got caught, stopped optimization, run it, and the test doesn't match... I don't know what's wrong. Maybe it's mine. I haven't contacted the service yet.

This has been seen even on mt4. Historical quotes (O,H,L,C) may change during optimization. Also the result of the most recent trade may change, if a new bar came during optimization. Also the spread may change: you started optimization with one spread and checked it with another. Although mt5 should include historical spreads (asks and bids), but I haven't checked it myself. Maybe knowledgeable people can answer about historical spreads in mt5. In mt4 the spread for the entire history was always taken equal to the most recent spread in quotes. Therefore, the test results varied depending on the time the tester was launched. Only on Saturday and Sunday the results did not change.

[Prival]

gpwr:
.... Maybe knowledgeable people can answer about history spreads in mt5. In mt4 history spread was always taken equal to the most recent spread in quotes. Therefore, the results varied depending on the time the tester was launched. Only on Saturday and Sunday the results did not change.

https://www.mql5.com/ru/forum/1031/page7/#comment_19983

now according to my observation this is being corrected. Filling the story with spreads

[Aleksey]

gpwr:
This has been seen even on mt4. Historical quotes (O,H,L,C) may change during pivoting. It may also change the result of the most recent trade if a new bar came during optimization. Also the spread may change: you started optimization with one spread and checked it with another. Although mt5 should include historical spreads (asks and bids), but I haven't checked it myself. Maybe knowledgeable people can answer about historical spreads in mt5. In mt4 the spread for the entire history was always taken equal to the latest spread in quotes. Therefore, the test results varied depending on the time the tester was launched. Only on Saturday and Sunday the results did not change.

Well, as for the spreads, it is well known. They have been added to the quotes a month ago (by cosmic standards :) and they have hardly changed. And the system itself does not expect such losses on spreads. Something is wrong here. Either I screwed up the code somewhere (which is unlikely), or the testing results wildly differ from those of optimization. This is already a bug. That's why I'm asking everyone - has it ever happened to anyone?