Not for MT developers! What to replace INIT_PARAMETERS_INCORRECT with ? - page 9
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
OK, you don't understand, but that's to be expected. For hearing people, I've already said all I wanted to say. I see no point in continuing.
And for me it is completely unexpected that anyone can see any difference between 1200 and 0012, if there and there function 1 is executed first and then function 2.
It's just amazing that there are so ... um... blind people...
why can't the first population only be built from "acceptable" parameters?
And by and large, what prevents subsequent populations from being collected only from valid parameters as well???The first question seems logical to me. It is possible to collect 512 valid individuals at the start of optimization. @Renat Fatkhullin, what's the problem with it?
But then, there's a problem. If a combination of the best of those 512 individuals gives you a non-valid result, by what right can you substitute another variant for it? Adopt a child if there is a miscarriage? This is no longer genetics.
The first question seems logical to me. It is possible to collect 512 valid individuals at the start of optimization. @Renat Fatkhullin, what's the problem with it?
Well, in the case of TC, genetics will not help. Maybe a little.
You can simply make an adaptive generation size, based on the number of valid passages in the first generation.
Well, in TC's case, genetics won't help at all. Maybe a little.
No, I'm not considering TC's problem. Genetics won't help her at all.
The first question seems logical to me. It is possible to collect 512 valid individuals at the start of optimization. @Renat Fatkhullin, what's the problem with it?
But then, there's a problem. If a combination of the best of those 512 individuals gives you a non-valid result, by what right can you substitute another variant for it? Adopt a child if there is a miscarriage? This is no longer genetics.
In my mind "a combination of the best of these 512 individuals" cannot give a non-faulty result. But after selecting the best individuals into the next population again to add individuals with acceptable parameters (weed out the invalid individuals) - what is the difficulty here?
To my mind "a combination of the best of these 512 individuals" cannot produce an invalid result.
Perhaps you should go beyond your narrow objective.
Perhaps you should go beyond your narrow task.
My "narrow" task is a special case. And, of course, I am quite interested in it ))))
But we are talking about a wider problem.
Here's another "narrow" example:
The strategy implies the use of a simple moving МА. But according to the strategy logic (it does not matter!) МА period must be in the range of 9-32, or 80-92.
How to exclude the range which is not necessary in optimization? The most obvious and easiest one:
Put the optimisation parameters:
Here we go again with a glitch in genetics due to 'unacceptable' parameters...
And why do you purposely include incorrect parameters in the optimization?
Incorrect parameters are generally disabled for 3 reasons:
1. the study of the area is not of interest (e.g. excluding a number of timeframes, as the most understandable option);
2. the area has been studied previously and there is no need to study it again;
3. the area does not exist. As an example, we can present a percentage oscillator where there are two levels - the upper and the lower; obviously, the upper level must not be higher than the lower one. Well and here I can tell you about mathematical mishaps, when a variable leads to division by zero and for this reason a certain combination is disabled.
Why would you intentionally include incorrect parameters in the optimisation?
Well, how do you organise the MA optimisation from the example without including the "incorrect" parameters?