EA N7S_AO_772012 - page 37

[Николай]

boing9267 писал(а) >>

N7S_AO_772012_L9_2, is there a more recent one?

There are a lot of versions, someone is remaking it for his own idea, someone uses the one I suggested. For the latter, I offer the following version N7S_AO_772012_M1.mq4 The former may use the following one.

In the latest version the main driving force is still the AO (Indctr = 1) and it is also used for optimization of inputs. However, there are some fixes.

First, I found an unnecessary check of the number of open positions in the function

double BTS() {

if ((prcptrnz1() > 0 || F==0))//& bu<HM_Up_X)

{if (prcptx1 > 0 && Delta_G12>0) {sl = slx; tp = tpx*slx; mn= mnx1; return (1);}}

if ((prcptrnz1() < 0 || F==0))//& sll<HM_Dn_Y)

{if (prcpty1 > 0 && Delta_G12<0) {sl = sly; tp = tpy*sly; mn= mny1 ;return (-1);}}

return (0);}

commented out. This resulted in that the reverse signal was not taken into account when positions were opened in opposite directions and closing "with the market" was not quite correct. This problem has now been solved.

There is one more "flaw" in the trl() function I recommend to delete the operator return(0); it will not be fatal, but still )))

The other changes are related to speeding up the optimization process, i.e.

void cmmnt() function doesn't interfere with optimization and in bool FLG (int cs ) it calculates hep parameters only for the first 14 hours.

Now the first and second steps (the longest ones) take about 35-40 minutes to optimize while it used to take over an hour.

[cross22]

Thank you :)

[[Deleted]]

Holy shit))))

Now let's put it on optimisation)))

[cross22]

What happens to the open position when you switch to the new version?

[Николай]

cross писал(а) >>
What happens to the open position when you switch to the new version?

Nothing catastrophic should happen. Trailing with the new stop levels will work and that's it, but the old sets are likely to work badly.

[Николай]

I'll post the sets for the new version in an hour and a half.

[Николай]

Presets on seven pairs for the current week. Optimized by 4+2 scheme. Frank is still poorly optimized, I have not found anything good for it yet.

Good luck and profit to everyone!

[Alexander Sevastyanov]

Vovanych >> :

The computer is a 4-core processor, 8GB of RAM.

MT4 tester cannot even use the 2nd core, let alone the 3rd and 4th.

So in case of optimization your engine is working at no more than 25% of its capacity. :(

Is it 8 gigs on server or 64 bit axis?

XP, vista etc address no more than 3 and a little gigs of OP.

[Евгений]

Thank you, SHOOTER777, for the work done. We will launch the new version next week. Ran it with the sets from L9_2. Of course there are differences, but only minor ones. The difference is about 5%, and in different directions, both better and worse. And the general character of the balance curve remains the same

[[Deleted]]

goldtrader писал(а) >>

The MT4 tester can't even use the 2nd core, let alone the 3rd and 4th.

So, when optimizing your stone doesn't work for more than 25% of its power. :(

Do you have 8 gigs on server or 64 bit axis?

XP, vista etc do not address more than 3 gigs or so of RAM.

8 is on vista 32, XP doesn't see that much memory))))

What a pity MT4 cannot do that yet)))) but I will not change the computer yet)))