MetaTrader 5 Strategy Tester: bugs, bugs, suggestions for improvement - page 32

 
fxsaber:

Then I try in the EA, which is running in the Tester, to get the original settings not through the ini file, but through the clipboard. But, unfortunately, I don't understand how to get the Terminal handle.

You can identify your own Terminal from the set of running Terminals by its path to Agent - it also contains path to Terminal. But getting a list of running Terminals with their paths to apply this clue doesn't work. If you tell me, it would be wonderful. Then we will be able to get the initial data of the Expert Advisor under test.

I have solved the problem.

 
Slava:
This was done intentionally, so as not to spoil the original set

Thank you. Is the following behaviour a flaw or not?


When selecting the optimisation cache, the Optimisation graph appears and the tester settings become equal to what is prescribed in the opt-file.

But parameters of advisor - no. In order to see, what were the parameters of advisor for optimization in the selected cache, you must necessarily select a single prog from the tab Optimization.

 

Please, if in the Optimisation tab a single run is applied, please disable Optimisation mode in the Tester Settings.

Now it is not switched off and you get a lot of erroneous Optimisation runs instead of single runs.

 
fxsaber:

It used to be very convenient to work with optimisation results. You highlight an interesting pass, then do sorting by some column, and the highlighted pass remains highlighted and visible.

Now, after sorting, just the top row is highlighted, the pass before that is completely lost from visibility.

Is it possible to bring back the old behaviour?

The specified behaviour is not reproduced. The highlighted row remains in the field of view, regardless of sorting

Exact reproduction steps needed

 
fxsaber:

Please, if in the Optimisation tab a single run is applied, please disable Optimisation mode in the Tester Settings.

Now it is not switched off, and you get a lot of erroneous Optimisation runs instead of single runs.

Where do erroneous Optimisation runs come from?

Do you not change the input parameters of the Expert Advisor for the next run?

 
Slava:

The specified behaviour is not reproduced. Highlighted row remains in the field of view, regardless of sorting

Need exact reproduction steps

I select previous cache recordings via the GUI. It always reproduces there.

 
Slava:

Where do erroneous optimisation runs come from?

Do you not change the input parameters of the Expert Advisor for the next run?

For example, I do the best run from the Optimisation tab. Then I want to see how that run would behave on a wider interval. I change the interval and don't always remember before clicking on Start that I need to turn off Optimization.


Also, after a run from the Optimization tab, I often change some input parameter to enable, for example, the generation of an HTML report at the end of a run. Then I press Start again, forgetting to turn off Optimization.


When I use Multitester, I get dozens of caches, which I review. And after choosing each of them, I have to switch off Optimization all the time, in order not to restart it by mistake.


Perhaps others have a different view, but my logic is this: once you start watching the results of Optimization, you are more likely to work with single runs, than to re-run Optimization.


So I suggest that if you select single runs in the Optimisation tab with the PCM, then disable the Optimisation mode.

 
Can you tell me how to clear the history of previous tests in the strategy tester?
 
fxsaber:

I select previous cache recordings via the GUI. It always plays there.

Are the previous caches genetic?
 
fxsaber:

Others may have a different view, but my logic is that once you start looking at the Optimisation results, you are more likely to work with single runs than to re-run Optimisation.

So I suggest that if the PCM selects a single run in the Optimization tab, then disable the Optimization mode.

Yes, I also run Optimisation all the time instead of a single run.

After running any single test, it is logical to exit optimisation mode.