Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Similar Threads

Looking for Trailing Stop EA like default MT4 trailing stop 7 replies

Trailing stop after up some pips and question trailing step 3 replies

Trailing step and trailing stop optimization 0 replies

Buy Stop and Sell Stop orders with OCO and Trailing Stop 0 replies

To Trailing Stop or Not To Trailing Stop.... 36 replies

Problem with Trailing Stop Conditional Checking

  • Post #1
  • Quote
  • First Post: Edited 4:32am Dec 25, 2009 2:25am | Edited 4:32am
I have an annoying problem with the conditional checking of my trailing stop feature, which causing error 130 (invalid stops)

Here is the condition which will call the SetTrailingStop function when it's returning TRUE (processed on BUY orders)

Quote
if(TrailingStop!=0 && NormalizeDouble(Bid,Digits)-NormalizeDouble(OrderOpenPrice(),Digits)>NormalizeDouble(Point*TrailingStop,Digits) && (NormalizeDouble(Bid,Digits)-NormalizeDouble(OrderStopLoss(),Digits)>NormalizeDouble(Point*TrailingStop,Digits))) SetTrailingStop(...);

I've also test it with the simple:
Quote
if(TrailingStop!=0 && Bid-OrderOpenPrice()>Point*TrailingStop && (Bid-OrderStopLoss()>Point*TrailingStop)) SetTrailingStop(...);

But both of them showing this result when i print it (it will only print when the result TRUE)
Quote
USDJPY BUY TS: 97.84000000-97.73000000(0.11000000)>0.10000000 & 97.84000000-97.74000000(0.10000000)>0.10000000
As you can see in the Bold, WHY 0.10>0.10 = TRUE ?? which make the EA calls the setTrailingStop function where it shouldn't, thus causing error 130 (invalid stops) because it tries to set the SL of a BUY order to the same price with the Bid.

So, anyone ever experience this kind of thing or know how to resolve it? (i wonder if it's a bug in MT4)
hey adam..... So, anyone ever experience this kind of thing ..... it'd be safe to say anyone that has ever typed for(int i=0; has experienced it.....

//----
or know how to resolve it? ....... reduce the line to it's simplist terms..... then check for logic errors.....

//----
(i wonder if it's a bug in MT4) .... those russian dudes are kinda smart..... i wonder what would be the odds of us discovering an error of theirs..... look within first.....

//----


......h






Inserted Code
   for(int i=0;i<OrdersTotal();i++)
     {
      if(OrderSelect(i,SELECT_BY_POS,MODE_TRADES)==false)        break;
      if(OrderSymbol() != Symbol()) continue;
      if(OrderType() == OP_BUY)
 
                  if(TrailingStop != 0 && Bid - OrderOpenPrice() > TrailingStop*Point && OrderStopLoss() < Bid - TrailingStop*Point)  
 
                    {
                     OrderModify(OrderTicket(),OrderOpenPrice(),Bid-Point*TrailingStop,OrderTakeProfit(),0,Red);
                     return(0);
                    }
      }
to trade and code, keep both simple... no call to impress....h
Well i don't think everyone will bump in the 0.100000>0.100000=TRUE

I've recheck everything all over and over, and as you can see in my 2nd quote it's the simplest condition checking of my code, nothing related with the FOR or anything, because the only problem occurs in the bold (the right most condition)

I had a bad feeling that the NormalizeDouble function didn't really cut the digits properly, i'll try making a similar function to see the result tomorrow (i've just got back from a very boring weekend and feels tired to do any programming today)

When i test it on a 5 digits broker i can see 0.101000>0.100000 which is correct.

Anyway, i've solve the "invalid stops" issue by making additional checking in the setTrailingStop function, but that 0.10000>0.10000=TRUE on 4 digits broker, is still making me curious, because this condition could also occurs in other parts of the code where i compare 2 double variables.
  • Post #4
  • Quote
  • Edited 9:04am Dec 28, 2009 8:33am | Edited 9:04am
I've found the problem O_o
Apparently the problem was because i use
&& NormalizeDouble(Bid,Digits)-NormalizeDouble(OrderStopLoss(),Digits)>
(which will give the same result as && Bid-OrderStopLoss()>), while the correct one should be
&& NormalizeDouble(Bid-OrderStopLoss(),Digits)>

After changing it, the 0.100000>0.100000 which resulting TRUE Never occur anymore because now i only see this one printed(when resulting TRUE) which is correct:
Quote
USDJPY BUY TS: 97.85000000-97.73000000(0.12000000)>0.10000000 & 97.85000000-97.74000000(0.11000000)>0.10000000

However, when i replaced the NormalizeDouble with my custom function:
Quote
double NormDbl(double val, int Dgt)
{
double pnt=1/MathPow(10,Dgt);
double tmp=MathRound(val/pnt);
return(tmp*pnt);
}
What ever condition format i use(using both ways including the one that show the bug), it always gives the correct result O_o i guess the NormalizeDouble does have something to do with the weird result.

So i will be using my custom function to prevent something like this happened in the future (somewhere in other part of the code where i use NormalizeDouble without cautions)
  • Post #5
  • Quote
  • Edited 3:07pm Dec 28, 2009 3:06pm | Edited 3:07pm
halo adam,

don't compare two double variables.
i tested the following code:
Inserted Code
double d1 = 97.84000000-97.74000000;
double d2 =  0.1;
if(d1 == d2) Print("d1:",d1," = d2:",d2);
else if(d1 > d2) Print("d1:",d1," > d2:",d2);
else Print("d1:",d1," < d2:",d2);
on journal tab:
Quote
2009.12.18 22:50 #EA0 GBPUSD,M30: d1:0.1 > d2:0.1
convert

to solve the problem, convert those double vars to integer first, and compare these integers:

Inserted Code
int i1 = MathPow(10,Digits)*d1;
int i2 = MathPow(10,Digits)*d2;
if(i1==i2) ...
should works
  • Post #6
  • Quote
  • Last Post: Dec 29, 2009 3:54am Dec 29, 2009 3:54am
Quoting sangmane
halo adam,

don't compare two double variables.
i tested the following code:
Inserted Code
double d1 = 97.84000000-97.74000000;
double d2 =  0.1;
if(d1 == d2) Print("d1:",d1," = d2:",d2);
else if(d1 > d2) Print("d1:",d1," > d2:",d2);
else Print("d1:",d1," < d2:",d2);
on journal tab:
convert

to solve the problem, convert those double vars to integer first, and compare these integers:

Inserted Code
int i1 = MathPow(10,Digits)*d1;
int i2 = MathPow(10,Digits)*d2;
if(i1==i2) ...
should works
Thanks for the test sangmane
I thought using NormalizeDouble could solve the problem, but apparently it can't solve it.

Yes, comparing 2 doubles will be better after it converted to interger, but it will looks less complex to use a custom function than converting it directly on each line
0 traders viewing now