Experts: Tester v_0.14

 

Tester v_0.14:

Example of using optimization reports in an Expert Advisor.

Author: George

 

This is so good.

It would be very useful if you describe a little about the strategy and the way you copied data from Optimizer.

Thanks,

 
ahpd:

This is so good.

It would be very useful if you describe a little about the strategy and the way you copied data from Optimizer.

Thanks,

Hi ahpd and everybody!

I'm gonna try to explain with my beginner English what you want to know.

There is no strategy needed for using this method. This is about using the optimizer of mt4 strategy tester to find the most perfect combinations, and then to let them confirm or cancel each other. You need to get intervals by rounding the values of the used indicators, then you can simply filt their positions with each other's one. You will need a simple code for this searching, which contains the just optimized indicators only - the more (even not used) indicators are there in the code, the more time the optimization takes. Buy and sell signals should be searched separately - otherwise one good buy signal will create different combination with every single good sell signal, and vice versa.

When optimization is finished, double click on "Profit factor" button on the top of optimization result - results will be ordered by profit factor. Then press right click on the results and select "copy all". Open a notepad, and paste them into it, save and close. I use openoffice calc to make code from them - open the saved file with that, select "tabulator" as separator. Here you can delete the unnecessary data, and replace the name of the extern variables with the name of the rounded values, and you can also paste the operators ( if, &&, ||, (), {, ;} ) with this replace function.

This one is an early version and gives this nice backtest result. I have found some more mistakes in it since i shared it:

- i should have used median prices only, because of the many difference in different broker's historical data (open/close values),

- some of the indicator value intervals are too short, giving only 1-2 result(s) in the whole history,

- many almost good combinations (between profit factor 1 and 2) was thrown away - they could have been numbered and filted again,

- it should have been tested on more broker's historical data.

I stopped working with this, because i get help from a professional programmer, and now we play together on smaller timeframes.

Good luck!

 
i just want to ask, how much money you recommend i need to use this EA ? Thank you.
Reason: