MetaTrader 5 Strategy Tester: bugs, bugs, suggestions for improvement - page 50
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
In the first screenshot you can see that there are completed tasks. There are 150 of them.
If you count the tasks in the job on the second screen, there are 103, of which 9 have been completed.
Well, we found 9 completed tasks. And where are the 150 completed tasks on the first screen?
Have you worked much with a strategy tester or do you want to give answers from pictures?
Have you worked much with a strategy tester or do you want to give answers from pictures?
In your questions, you operate to the screenshots provided. Therefore questions on the pictures. So where on the first screen shows that the completed tasks 150?
I addressed the developers in a dedicated thread. Please do not disturb sofa analysts.
I addressed the developers in a dedicated thread. Please do not disturb sofa analysts.
There are already 150 results and it shows 9. Why?
This is the picture in the agents:
You can't tell anything from these pictures. You have to look at the logs. Full logs are better.
Where is the left side of the optimisation results tab?
I can assume that the first 141 results came from the optimization cache, which was previously run with the same parameters
I suggest considering several requests for the tester report:
1) highlight the initial deposit on the balance chart with a solid line (moments above/below the initial balance will be visible). both in the report and in the tester itself.
2) now it is difficult to check whether the values of the current pass correspond to the previously made report, it would be easier if the values in the report and what I see on the screen coincide:
- display variables of datetime type as time, not seconds since '70.
- the same as in the previous sub-item regarding colours (in the report we see 8421504 and in the initial menu Gray), timeframes... as a program maximum, all enum, because it is difficult to make settings like in the report.
- optionally specify a variable description (the variable name is not visible to the user, only the description, but the opposite is true in the report)
3) almost from the realm of fantasy, but quite feasible: if it is possible to load parameters not only from a set file, but also from a report (xml and htm). they are there almost in the same form as in the set file! then point 2 will be 80% irrelevant
You can't tell anything from these pictures. You have to look at the logs. Full logs are better.
Where is the left-hand side of the optimisation results tab?
I can assume that the first 141 results were loaded from the optimization cache, which was run before with the same parameters
By the way yes, was re-run. The pair was very slow to optimise.
Maybe then it makes sense to write not 9/1280 but
150/1280 or 9/1130 ? (better the first option). Then re-run will show that results are taken from cache, because 1280 is general number for my runs, regardless from cache or not.
I recompiled the EA while it was running to check - the optimization didn't stop, everything is OK. But after restarting, results seem to be taken from cache, but duplication has started.
I have not seen any difference between the two EAs, but I have not seen any difference between them.
Regarding the account (written before) - if you switch, the optimisation stops.
This is good, because the data from servers of different accounts may not match. For example, the tester didn't have enough history, so he downloaded it from the current server and got a chimera.