How do I check whether 'Optimisation' or 'Forward Optimisation' is in progress? - page 11

 
Youri Tarshecki:
The individual date will of course be visible, but you won't be able to tell by the individual month.
You can move the mouse left and right to find the month boundaries. Well, the dates are signed at the bottom. The standard terminal does not provide any other.
 
elibrarius:
You can move the mouse right - left and find the month boundaries. The dates are signed at the bottom. The standard terminal does not provide any other.

And how do you extract monthly statistics from such a graph? The in-house tester does it better for both back and forward, so it's easier to stop every month to dump the results and pictures into a file and move on.

In addition, accounting for previous balance is only harmful for evaluating monthly performance.

 
Youri Tarshecki:

And how do you extract monthly statistics from such a graph? The in-house tester does it better for both back and forward, so it's easier to stop every month to dump the results and pictures into a file and move on.

I do all this before... First, I perform both back and forward tests with standard tester, stop and look at results, save them in a file. And then I run it in swapped input mode.

Also, counting previous balance is only detrimental to monthly performance evaluation.

I can see monthly performance through the results of forward testing. Switching input data method allows to imitate real trading. Suppose you started trading on Jan 1, 2015, one month later you overoptimized and changed your EA's inputs and continue trading and so on every month.

Or do you want to close a working account every month, withdraw money and open a new one? Or alternatively, close all trades, withdraw money before the initial deposit and start trading from scratch? If so, a standard forward pass is enough for you.

By the way, you can simply overlay the charts in Photoshop, if you can't build a script.

 
elibrarius:

I do all this before... First, I perform both back and forward tests using the standard tester, then I stop and look at the results, save them to a file. And then I run it in swap input mode.

Monthly performance can be seen on the basis of forward testing results. Switching input data method allows to imitate real trading. Suppose you started trading on Jan 1, 2015, one month later you overoptimized and changed your EA's inputs and continue trading and so on every month.

Or do you want to close a working account every month, withdraw money and open a new one? Or alternatively, do you want to close all trades, withdraw the money before the initial deposit and start trading from scratch? If so, then standard forward pass is enough for you.

By the way, you can simply overlay the charts in Photoshop, if you can't build a script.

The wolfing forward is a method of testing an EA's behavior on a non-optimized environment, not the behavior of a trader in terms of taking/withdrawing profits.

During a single back-and-forth run, exactly the same starting balance is loaded into the tester for the back and forward ones. And this is correct. Therefore, you can easily compare the performance of the EA with the back end as well as with other forwards in other segments or other EAs in general. Unequal starting balances will not change the performance on a particular month at all, because it is a relative value, but its estimation will be difficult.

That's why, by the way, all tests should be performed on a fixed lot.

Reason: