Cool system! - page 35

New comment
 
It's reliable, but a separate site could be set up, so think and write your wishes.
 
artsnz >> :

>> If everything is so good, why don't you share your EA or the strategy it is made with?


I didn't claim that my EA is perfect, so I don't see the point in posting it.

I don't know how it works and change settings blindly (of course there are settings, but they were previously set inside the Expert Advisor).

I don't understand how it works and I cannot change its settings blindly (of course they are available, but they are already preset inside the Expert Advisor)

If you personally are interested in it, I can share. on this matter in a private message.

It works relatively well, but ... but it's a rare and accurate loser, 1 or 2 orders, so...

great as 99% of other EAs, i'm just trying to get the EA to be minimally dependent on the market.

The problem is not putting the right orders, the problem is fighting the wrong orders.

Even if I have 1 losing order and 9 profitable ones, this 1 has to be closed correctly to stay in profit.

In my system, the number of losing orders is much less than the number of profitable ones but taking into account the large distance of losses, we can let them all go. I try to dynamically calculate the distance to a stop loss depending on the market situation.

The strategy is as follows:

IBFX - we take a buy or sell command and take profit.

Confirmation of direction - Dynamic extrapolator. ( https://www.mql5.com/ru/code/9121 ) author : neoclassic

The extrapolator has been finalized by the author at my request, the goal is to calculate when requested, not once (I have to thank the author very much for that).

Checking the extrapolator on H1 and M15

The extrapolator is requested only when a bar changes or by a condition (otherwise the calculation on each tick will practically hang the EA)

If everything is the same, a pending order is placed at a given distance.

If the order has missed the wrong side and the pending order did not trigger, the pending order is placed at the given distance.

Stops are placed at a large distance, because the system is not perfect, and there should be space for drawdowns and reversals.

When the order triggers and after the price goes in the plus direction, stop loss is set at a distance behind the price, for insurance, if it does not reach Take Profit.

Lot is set dynamically - percentage of the deposit is specified.

The problem with the strategy, dynamically changing stop losses, depending on ....... who knows what.

So, when they are not needed they should be far away, when they are needed they should work out with a minimum of loss.



 
I'm still new to all this and just learning, but about the site I can help you create a private forum, hosting - site available.
 
artsnz >> :
But it's reliable, but we can make a separate site, so think and write your wishes.


Regarding your EA.

There is a suggestion to speed up the work.

Only call the indicators when the bar changes.

In any case, indicators do not take data from ticks, but from existing bars up to the last closed bar.

And they don't care about every tick.

example (IMHO):

//-----------------------

int bars=GlobalVariableGet("bars"); 
 if ( bars != iClose(NULL,0,1)) 

{

GlobalVariableSet("bars", iClose(NULL,0,1));

Индикаторы

Индикаторы

Индикаторы

}
 //-------------------

After this they will be called only after a bar change.

Indicator data can be read from buffers, imho, it will also affect acceleration.

 
The deal has been on the chiff since this morning... Was at -35... now at +20... м.. the trawl in the code was a 15... everything's fine... not working? I don't quite understand about opening a trade... There have to be 3 conditions - shape, directionality from x1 to day (or x4) and blue-green-yellow-red line... Yellow gives a 28% probability... It seems that the advisor does not take into account which line and consequently what is the probability of trend direction...
 
ixoid >> :
I am a newbie in all this and just learning, but about the site, I can help you create a closed forum, hosting - site available.

What's the point? You won't get any new people or ideas.
I think it's pretty good.

 
I haven't had an advisor in two days, it's boring :|
  
static double LineGreen, LineRed;
static datetime dtH4 = 0;
if ( dtH4 != iTime(NULL, 240, 0))
{
  dtH4 = iTime(NULL, 240, 0);
  // TRO_InsideBar_Plot2.
  if (IsTesting()) iCustom(NULL, 240, "TRO_InsideBar_Plot2", 50, true, Blue, Lime, 0, 0);
  LineGreen = ObjectGet("HIGH_0", OBJPROP_PRICE1);
  LineRed = ObjectGet("LOW_0", OBJPROP_PRICE1);

  // IBFX - CPR.
  if (IsTesting()) iCustom(NULL, 240, "IBFX - CPR", true,  true, true, true, DarkBlue, DimGray, 
                                                            Lime, Red, Blue, FireBrick, 50, 0, 0);
...
}
Many variables are sufficient to recalculate once per bar (4H, 1H, M30) and not on every tick. Example above. Am I thinking correctly?
 
I paid attention to the IBFX parameters GMT change
Depending on this parameter the signals can change drastically

In general GMT should be set to the time of the server of the brokerage company.

 
Alivio >> :
I haven't had an advisor in two days, it's boring :|


The signals will be really scarce. I have to ask the topicstarter about the frequency of trades.
1...282930313233343536373839404142...48
New comment