Your symbols and your datafeeds in Metatrader 5 - page 3
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
2015.04.22 19:44:18 Tester genetic mode turned on due to 567869252041 passes, to use complete algorithm decrease amount of optimised parameters
I need full brute force, what to do?
567 billion passes even if you count per 100ms pass, it's still 56 billion seconds.
Are you sure you want to wait 648,000 days (1,775 years) or take advice to switch to genetics? With genetics you will shoot off in 20,000 passes and you will be happy.
Renat, by datafeeds you mean ticks or bars, i.e. the tester will work on real (recorded ticks) and not on generated (simulated) ones?
Will it be possible to record ticks and bars by itself, or to download them from another source and then test using them?
The management of both ticks and bars and glasses is available.
If, of course, the particular datafeed supports it.
567 billion passes even if you count per 100ms pass, you still get 56 billion seconds.
Are you sure you want to wait 648,000 days (1,775 years) or will you take the advice to switch to genetics? On genetics, you will shoot off in 20 000 passes and you will be happy.
is this a hardware problem or a platform problem?
More like your awesome mind. You cannot turn off your brain and disconnect from the physical meaning of the processes.
Just to break through the physical boundaries, heuristic methods are invented to reduce by many orders of magnitude the complexity of calculating the result of combinatorial explosion. One very effective method is genetic modelling of a set of modifiable parameters for searching/finding work areas.
The terminal deliberately suggests switching to genetics because it realises that there is physically no chance of achieving a result by blunt brute force.
This particular task of yours, spread across agents, will require an extreme amount of financial resources.
It is likely that even a couple of generations of your descendants will not be able to pay it off.
So rather learn how to use a calculator. It's a primitive task, after all.
More like your awesome mind. You cannot turn off your brain and disconnect from the physical meaning of the processes.
Just to break through the physical boundaries, heuristic methods are invented to reduce by many orders of magnitude the complexity of calculating the result of combinatorial explosion. One very effective method is genetic modelling of a set of modifiable parameters for searching/finding work areas.
The terminal deliberately suggests switching to genetics because it realises that there is physically no chance of achieving a result by blunt brute force.
It is good that you have started to suspect something.
I recommend everyone to read the documentation and related articles - this will help you understand how things work and answer hundreds of questions.
is this a hardware problem or a platform problem? what hardware is this calculation for?
When there are too many variables to optimise, there is a high probability of fitting to history.
As variables grow, optimization time grows exponentially and there's no getting away from it.
There is a way out of this, a super powerful computer the size of a house)
Renat:
It is precisely to break through the physical boundaries that heuristic methods are devised, reducing by many orders of magnitude the complexity of calculating the result of a combinatorial explosion. One very effective method is genetic modelling of a set of modifiable parameters to find/find workable domains.
The terminal deliberately suggests switching to genetics because it realises that there is physically no chance of achieving a result by blunt brute force.
There are heuristic methods that are best suited to optimizing trading systems. They are customized for this very purpose, unlike the universal GA. Tell me, is it possible to replace the only heuristic method in the Strategy Tester with your own one written in MQL5? If we are going to dramatically improve the tester in terms of customization of the input data, it would be reasonable to customize the optimization methods (custom optimization criteria have long been available) to give it the best performance.
Pay attention to this topic on tester and optimizer acceleration. There's a technical example of feasibility there in addition to the proposal.