I'm trying to transition to SQX and I realize the huge effort the team has been placing in making the tool a real usable and stable solution at the same robustness level as SQ3. But every time I decide to try a definititve transition there has been always some malfunctions that really discourage me to go on as they are time consuming, on top of the learning curve I already have to invest to make the transition to a more complex and complete tool.
Here's my latest experience:
I'm trying to develop a H4 USDCAD strategy. To warrant to find enough strategies in a reasonable time I've used (what I think to be) a very basic config file (in attach).
Upon developing, I found the following problems:
a) The builder stopped after it had found 2055 strategies with the message Databank full - I had set the databank to hold 5.000 strategies and to "never" stop. See images 1 and 2
b) when I click in some of the strategies in the databank/results list, it returns the error "strategy doesn't exist" error. See image 3
c) The best strategies found in Builder are nowhere to be seen in Results/Databank list. See image 4.
d) There are problematic strategies that appear in the results list. Probably this is not an error, but by default I think the list should just eliminate these strategies. There is already too much data to handle to even one have to be stumbling in these strategies. If certain users want to have them, then they could chose by some checkbox or something. I believe most traders will not want to lose time with strategies that are problematic at the start of a long workflow.
e) When BT what it seems to be one of the best strats that one can access, the EA results in many trading errors in Dukascopy tick data. Because this might have a totally different problem, I open another ticket for this problem and will make a reference to this ticker in it.
Description changed:
I'm trying to transition to SQX and I realize the huge effort the team has been placing in making the tool a real usable and stable solution at the same robustness level as SQ3. But every time I decide to try a definititve transition there has been always some malfunctions that really discourage me to go on as they are time consuming, on top of the learning curve I already have to invest to make the transition to a more complex and complete tool.
Here's my latest experience:
I'm trying to develop a H4 USDCAD strategy. To warrant to find enough strategies in a reasonable time I've used (what I think to be) a very basic config file (in attach).
Upon developing, I found the following problems:
a) The builder stopped after it had found 2055 strategies with the message Databank full - I had set the databank to hold 5.000 strategies and to "never" stop. See images 1 and 2
b) when I click in some of the strategies in the databank/results list, it returns the error "strategy doesn't exist" error. See image 3
c) The best strategies found in Builder are nowhere to be seen in Results/Databank list. See image 4.
d) There are problematic strategies that appear in the results list. Probably this is not an error, but by default I think the list should just eliminate these strategies. There is already too much data to handle to even one have to be stumbling in these strategies. If certain users want to have them, then they could chose by some checkbox or something. I believe most traders will not want to lose time with strategies that are problematic at the start of a long workflow.
e) When BT what it seems to be one of the best strats that one can access, the EA results in many trading errors in Dukascopy tick data. Because this might have a totally different problem, I open another ticket for this problem and will make a reference to this ticker in it.
Attachment 3 Strat_exists_not.jpg added
Attachment 4 BestNotExist.jpg added
Attachment 5 Problematic_strats.jpg added
Attachment 20190702_SQX_Settings.cfx added
Attachment 7No_Save.jpg added
Need to add an extra problem.
d) When i tried to save the strategies to at least retain some of the time I've spent, SQX doesn't allow me to do so (Saving reports failed). See image 7.
So it seems the only option I have is to restart the whole process and hope that no errors occur in the next instance.
Attachment screen.png added
thank you for your report.
The problems a, b and c should be fixed in the new upcoming release. I've fixed a few bugs related to databank recently which were causing this kind of behaviour.
d) you can set up automatic dismissal of problematic strategies in Ranking settings - see the attached screenshot
Regarding the saving error - can you attach also your log file? It is located in user/log/StrategyQuant/log_{date}.log
Best regards,
Tomas