Printable View
Hello everyone, I have been working with a developer to create an expert advisor and across the line, I got puzzled on why it arouses some good trades. In addition, it fails to exit some transactions once an opposite signal comes up, though it was programmed to do so. Can anybody assist me with good suggestions as to get to the bottom of this? Lots of thanks in advance for your ideas.
Seriously, you can not think why? Face palm...
Quote:
Originally Posted by ;
If you post the code up, I am sure someone will probably be more than willing to take a look, but without the code no one will have the ability to answer your question.
It is the difference between your subjectivity, your unconscious urge to see just good trades and fantastic exits after the charts are already attracted, and the objectivity of this EA that just follow the principles you gave him without a subjectivity. It is the difference between losing and winning. Incidentally, if you can not trade manually profitably with the rules of your own system, don't bother with an EA. It will not do better than you. If you are not patient enough, then trade at least a commerce simulator instead of backtesting. You'd be surprised.
Wtf is that. . I dont get it
If by lacking great trades you imply that your EA is seemingly not following the rules you think you have programmed into it then the issue is faulty signal. The next step is to look at your logs and trading diary for order errors and if there are none then your logic would be to blame.
The EAs that I write have a subroutine (procedure) that dumps the values of critical registers (variables) to a document, which I call from egic points in the code, even if a'debug' switch is turned on. This makes it possible for me to assess what these values were, therefore I can later review precisely why my EA took transactions that it should not have, missed transactions that it should have obtained, or otherwise does anything else that seems incorrect, by walking through the code; a step beyond what the data in MT4's journals and logs have to offer. Once I am satisfied that my code is working 100% correctly, I simply set the debug switch to'away', and I periodically prune or delete the dump files (as they tend to get quite large).Quote:
Originally Posted by ;
Some of my EAs are complex, and without this facility, I would see them extremely tough to debug. Historical scenarios in the market can't easily be reproduced, and I would otherwise have no way of knowing exactly why my code failed to deliver the desired outcome.
My EAs also have a centre to take https://docs.mql4.com/chart_operations/windowscreenshot of this chart when transactions are entered or left handed, which can also be helpful for reviewing historical scenarios.
I also write EAs for some of my trading buddies. Reviewing the procedure in the above mentioned manner helps to determine whether the code is faulty, whether they've neglected to explain their trading principles to me correctly, or whether there's some other motive (e.g. an unexpected spike in the spread), for any anomalies that they allege have happened. I normally walk through the debug process together on skype, which helps to give them an appreciation of how EAs operate.
Maybe you could discuss this type of egy with your developer. I have attached a code fragment below (the debug subroutines, replied from one of my EAs) for instance for him/her.
https://www.nigeriaforextrading.com/...8971816957.mq4
It's a Fantastic question