Discussion of article "Moving Mini-Max: a New Indicator for Technical Analysis and Its Implementation in MQL5"

[MetaQuotes]

New article Moving Mini-Max: a New Indicator for Technical Analysis and Its Implementation in MQL5 is published:

In the following article I am describing a process of implementing Moving Mini-Max indicator based on a paper by Z.G.Silagadze 'Moving Mini-max: a new indicator for technical analysis'. The idea of the indicator is based on simulation of quantum tunneling phenomena, proposed by G. Gamov in the theory of alpha decay.

Author: investeo

[Franco]

Awesome idea I really like it, going to do some backtests with this!

Good job! 

[Jin]

It's a very interesting article. Thanks to the author.

[NFTrader]

love it!

[NFTrader]

But how one can back-test it since it's drawings is only for a limited timespan (for the last recent bars only), I tried to edit the code so it can draw for all bars available without me messing, but I haven't succeed! 

am I missing something? 

[Kebin Du]

I love this article very much. I will modify it and use it for my daily trade.

[muller]

Very good article, thanks.

But i've played with it on forward testing:

- it badly repaint

- it only lately confirm the past

- the  lastest data are not drawn: looks like a sift of 4 or bars

So as it works at time, there is to many issues.

[conanlh903]

I really apprecialte this article.

However, I have read the original paper and your code. I have a question about the code following:

 for(k=0; k<m; k++)
        {
         sqiip1 += MathExp(2*(S[m-1+i+k]-S[i])/(S[m-1+i+k]+S[i]));
         sqiim1 += MathExp(2*(S[m-1+i-k]-S[i])/(S[m-1+i-k]+S[i]));

         dqiip1 += MathExp(-2*(S[m-1+i+k]-S[i])/(S[m-1+i+k]+S[i]));
         dqiim1 += MathExp(-2*(S[m-1+i-k]-S[i])/(S[m-1+i-k]+S[i]));       
        }

this is derive from this formula

since the footnote of i in formula was changed by m-1+i in the code, why the other part of the code does not change the footnote of i?

I mean: shouldnt this code would be like following?

sqiip1 += MathExp(2*(S[m-1+i+k]-S[m-1+i])/(S[m-1+i+k]+S[m-1+i]));

[conanlh903]

Another question I want to raise about this factor is when I caculate Qi,I use the data in the next several bar...so when I move the total timewindow I should recaculate all the Usi from the begining to the ending, the Usi for the same price point would not be the same which make the back-test make no sense.

[teguh joko widodo]

i think this indicator just for meta5. how could i attach it at meta4? or am i missing something? thanks in advance.

regards.

[Sergey Golubev]

Yes, this is for MT5. But indicators for MT5 can not be used in MT4; and indicators for MT4 can not be used for MT5. At least - it is what I know for now. For the future - no idea : some people say the we will have one MetaEditor for MT4 and MT5 and so on so I hope that it may be changed something with it.