The Sultonov system indicator - page 113

New comment
 
Yousufkhodja Sultonov:

Describe the gradient methods, please, briefly or cite sources for the most complete description of the method.

In brief: Gradient methods.

You can read more about it here.

There are also references to the literature.

 

.

 

.

 

.

 

.

 
Yousufkhodja Sultonov:

Then, follow Semko's recommendations and prophecies, I am not imposing any TC on you.

Yusuf, there is no need to make things up. I have not made any recommendations or prophesied anything.

All I tried to do was to save your time so that you would not waste it on useless bullshit. And also, so that you don't waste others' time with this crap.

Yousufkhodja Sultonov:

It was explained to you, it's something else entirely. It's like, "I'm looking for this, I don't know what!"

I had almost no doubt you'd get caught up in the lack of a loose dick. And you didn't realise that it was you who was being mocked by Fedoseyev )))

But, admittedly, I was hoping that you were more reasonable than you seem and understood that an SLAU with and without a free member is essentially the same thing - the same eggs, but in profile. And a free member will not add order to this chaos.

Think about it.
Here is an example of an ordinary SLAU:

x0 = a1*x1 + a2*x2 + a3*x3

x1 = a1*x2 + a2*x3 + a3*x4 

x2 = a1*x3 + a2*x4 + a3*x5

But with a free term.

x0 = a0 + a1*x1 + a2*x2 + a3*x3

x1 = a0 + a1*x2 + a2*x3 + a3*x4 

x2 = a0 + a1*x3 + a2*x4 + a3*x5

...after adding up the three equations...

we get:

a0 = S0 - a1*S1 - a2*S2 - a3*S3

где S0..S3     - среднее арифметическое столбцов матрицы
S0 = (x0+x1+x2)/3
S1 = (x1+x2+x3)/3
S2 = (x2+x3+x4)/3
S3 = (x3+x4+x5)/3

after substitution we again get an SLAU without a free term.

∆x0 = a1*∆x1 + a2*∆x2 + a3*∆x3

∆x1 = a1*∆x2 + a2*∆x3 + a3*∆x4 

∆x2 = a1*∆x3 + a2*∆x4 + a3*∆x5

где ∆x0..∆x5  - дельта (приращение) цены к среднему арифметическому текущего столбца.

где ∆x0 = x0-S0 = x0-(x0+x1+x2)/3 и т.д.

I.e. if in the first case was an SLAU of cents, in the case of SLAU with a free term a0 will be transformed into a simple SLAU of increments.

7th grade maths.

So what difference does it make!

If an attempt to use SLAU for price analysis and prediction is just stupid, then the input of the free term a0 is stupidity squared.

And no matter how you pompously call this nonsense ("Let's compensate our digression by introducing the concept of C0 - accounting for the pressure of historical data on the price at the beginning of analysis, assuming that the market has a memory. ") - stupidity will still be stupidity.


Moreover, my indicator is very easy to remake for your SLAU with a free term A0. You only need to add a few lines of code without changing the main function SLAU()

There will be the same white noise and the same yield as in the prediction of price movement using a coin.

But I won't be lazy. I'll spend another 30 minutes and do it.

 

Kolya, I will show your delusion with a simple example, after which you will be forced to admit that, in such cases, the presence of a free member is absolutely necessary. Let's look at the examples you gave:

Here's a simple SLAU for an example:

x0 = a1*x1 + a2*x2 + a3*x3

x1 = a1*x2 + a2*x3 + a3*x4 

x2 = a1*x3 + a2*x4 + a3*x5

Если представить, что, между х-сами нет никакой зависимомти, то, а1=0; а2=0, а3=0 и получим, что и х0=0, х1=0 и х3=0! Получили нулевые расчетные значения х-сов. Это нонсенс.

but with a free term

x0 = a0 + a1*x1 + a2*x2 + a3*x3

x1 = a0 + a1*x2 + a2*x3 + a3*x4 

x2 = a0 + a1*x3 + a2*x4 + a3*x5
В этом случае, получим х1=а0, х2=а0, х3=а0, что указывет, всего навсего, на отсутствие зависимости между х-сами, без каких-либо парадоксов. Тепеь, поняли своё заблуждение?
 
Nikolai Semko:

Yusuf, don't make things up. I have not made any recommendations and I have not prophesied anything.

All I was trying to do was save you time, so that you would not waste it on useless bullshit. And also, so that you don't waste others' time with that crap.

I had almost no doubt you'd get caught up in the lack of a loose dick. And you didn't understand that it was you who was being mocked by Fedoseyev )))

But, admittedly, I was hoping that you were more sensible than you seem and understood that an SLAU with and without a free member is essentially the same thing - the same eggs, but in profile. And a free member will not add order to this chaos.

Consider this.
Here is a simple SLAU for example:

But with a free term.

...after adding up the three equations...

we get:

after substitution we again get an SLAU without a free term.

I.e. if in the first case was an SLAU of cents, in the case of SLAU with a free term a0 will be transformed into a simple SLAU of increments.

7th grade maths.

So what difference does it make!

If an attempt to use SLAU for price analysis and prediction is just stupid, then the input of the free term a0 is stupidity squared.

And no matter how you call this nonsense high-sounding ("Let's compensate our digression by introducing the term C0 - accounting of pressure of historical data on the price at the beginning of analysis, supposing that the market has memory. ") - stupidity will still be stupidity.


Moreover, my indicator is very easy to remake for your SLAU with a free term A0. You only need to add a few lines of code without changing the main function SLAU()

There will be the same white noise and the same yield as in the prediction of price movement using a coin.

But I won't be lazy. I'll spend another 30 minutes and do it.

All in all -- nonsense.

On points:

1) this silliness does not come from great intelligence, but from a lack of knowledge and understanding;

2) lack of understanding of the role of the free member leads to statements like this : "that SLAU with a free term and without it are essentially the same";

3) to solve the problem it is not necessary to do: "after adding three equations";

4) this is one more indirect confirmation of elimination of trend component on increments: "after substitution we again obtain SLAE without a free term";

5) for solution of optimization problems "7th grade mathematics" is obviously not enough, so one should broaden one's horizons;

6) exclamations "stupidity","stupidity squared" and other variations with "stupidity" - this is from lack of knowledge and understanding (see item 1);

7) instead of reconstructing of an indicator from a problem of the solution of a system of linear algebraic equations, you should make one more indicator with the solution of an optimization problem, and then compare their readings, and as a result see and understand how different their readings will be, i.e. the solutions and statements of problems.

 

Here's your infamous a0 (aka C0)

White noise is white noise in Africa


I have a feeling that you gave birth to SLAU of 5 equations for years. And you've been dubbing it with a halo of mega-scientific sensation and delusions of grandeur. And that's 7th grade high school maths.

But my tiny SLAU() function easily solves SLAU of 50 equations and I made it and debugged it in less than 1 day. I don't know which way I solved SLAU, because I'm always too lazy to study existing methods, it's easier to invent my own. Most likely my way is not optimal and of course I haven't invented anything new, I'm not strong in theory. But I haven't come across a more compact one.

void SLAU(double &x[],double &f[],double &a[],int m)
  {
   int k=m-1;
   if(m>1)
     {
      double xx[],ff[];
      double g=x[0]; if(g==0) g=1.0 e-100;
      for(int i=0;i<ArraySize(x);i++) x[i]/=g;
      for(int i=0;i<ArraySize(f);i++) f[i]/=g;
      ArrayResize(ff,k);
      ArrayResize(xx,k*k);
      for(int i=0; i<k; i++)
        {
         ff[i]=f[0]*x[(i+1)*m]-f[i+1]*x[0];
         for(int j=0;j<k;j++) xx[i*k+j]=x[j+1]*x[(i+1)*m]-x[(i+1)*m+j+1]*x[0];
        }
      int i=0;
      for(;i<k; i++) if(xx[i*k]!=0) break;
      if(i>0 && i<k) for(int j=0;j<k;j++) {double t=xx[j]; xx[j]=xx[i*k+j]; xx[i*k+j]=t;}
      SLAU(xx,ff,a,k);
     }
   double sum=0;
   for(int i=1; i<m;i++) sum+=a[n-m+i]*x[i];
   if(x[0]!=0 && x[0]==x[0]) a[n-m]=(f[0]-sum)/x[0]; else a[n-m]=1.0/n;
   if(m!=n) return;
  }
Files:
SLAUPlus.mq5  28 kb
 
Олег avtomat:

Instead of redesigning the indicator from the problem of solving a system of linear algebraic equations, you should make another indicator with the solution of the optimization problem, and then compare their readings, and as a result see and understand how different their readings will be, i.e. solutions and problem statements.

So do it, if you have to. Why are you making all this fuss...

Everyone knows how to copy wiki...

ZZZ went to your profile.
All your ranking is based on comments. Not a single code.
Without a code your opinion is not authoritative for me.
Keep commenting...
1...106107108109110111112113114115116117118119
New comment