Optimise an EA and get the best of the optimised ones. - page 16
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
No, of course not (for fuck's sake, let's be on a first-name basis)...
You've seen my charts and tables - they're all done in Excel.
I have scripts that analyze the history and output a CSV file that can be easily imported into Excel.
I'll have to write another script which will output a similar CSV file with notes on what's done and what isn't.
Actually, if you're not afraid of the words MQL, OOP,"virtual functions", "virtual interfaces" - I can even show you the source code.
I don't understand where the script will take the data from, what has been done and what hasn't?
I'm not interested in the source code, I'm interested in the idea.
I have optimized another 6 EAs using the same currency.
Tough job dragging the behemoth out of the swamp
In general, the tests and real time is different
This is the first one.
Closer to business are EAs based on standard TA
This is the second
At the end of the day there is a problem that after the test, the Expert Advisor will be profitable for the next few days
This is third
In general, I don't understand where the script will take the data from, what is already done and what isn't?
I'm not interested in the source code, I'm interested in the idea.
The script takes the data on TC that needs to be reoptimized, plus it adds TC that do not exist yet.
Look for yourself - there are 16 Expert Advisors per symbol. Now the system is "familiar" with 25 symbols.
Accordingly, the script runs through the existing Expert Advisors and looks whether they exist, and if they do - how they trade, whether they have not stopped.
Further on, we get an excel spreadsheet containing names of Expert Advisors that do not exist yet and need to be reoptimized.
Optimised the following 6 more EAs, on the same currency.
I will process, invest, you will have 6 more reğods in your account.
At the weekend, I put seven TCs on real.
I thought they would go into the woods, and it would be good if they didn't go into a slump.
But only one trade showed a loss, generating three SLs (up to 11 SLs in a row are allowed).
The rest all went up, and so amicably that I even got scared. This was not good.
Alexei - private message to you with the investment password, you can have a look at this mess.A few advisers are optimised and so on
It's a tough job dragging a behemoth out of a swamp
In general, the tests and real time are different
This is the first.
And let them differ. There have been tests, "training", so to speak, on which all have performed well. Now it's "qualifying games in the League".
Let's see what they show.
I have over 250 TCs in the League, all based on different principles. And the League is remarkably similar to any human collective.
Half of it is a "swamp" working "purely for a paycheck". They earn their spread, nothing more.
A third are a "headache for the manager" - immediately after "going to work" they start drinking, conflicts, marriages...
But, about 10 per cent are "locomotives", who are bursting at the top and showing good results NOW. But tomorrow they may turn into a "swamp", or even a "headache"? They just might ! But for now, they are doing well, and we must take advantage of that.
Closer to the point are Expert Advisors based on standard TA
This is the second one.
At the end of the day there is a problem that after the test the Expert Advisor will be profitable in the next few days
This is the third
The principles on which the League's EAs work are the simplest and most "dumb", I have described them and they are also available in the description included in the archive on the Yandex CD. They are time-tested, and no one doubts that they exist. The only thing people doubt is their ability to work. I myself have doubts and see that in most cases they do not have any "effect". But, in a small number of cases there is a payoff - these are the cases I am trying to use.
And about "will be profitable later" - I think this is a gross mistake. We have conducted a one year test (with a 7 month forward). On this area this TS has shown some results, and the "tolerance limits". While the TS is within these limits - it shows the same behavior, and we think it is stable. But as soon as the TS leaves these limits - this is a clear signal that the TS has lost its stability. Using such TS is not allowed.
And there's even an opinion that we should set TC limits not only by losses, but by profits as well. And we should remove from trading not only those systems that show abnormal losses, but also those that show abnormal gains. And I am inclined to agree with that, although "cutting the hen that lays the golden eggs" will not be easy.
Here, yesterday there were outsiders to be coached
(Period 27.03.17 - 27.03.18, forward starting 27.08.17):
Symbol System
EURAUD EMAFlatRTS
NZDJPY EMATrendSAR
EURUSD ChnFlatDTS
CADJPY EMATrendDTS
AUDUSD EMATrendSAR
EURCAD EMAFlatSAR
EURNZD EMATrendDTS
EURNZD EMATrendSAR
GBPUSD EMAFlatDTS
GBPJPY ChnTrendRTS