Sultonov differential indicator - page 42

[Dmitry Fedoseev]

Yousufkhodja Sultonov:
I meant this: "Keep in mind that RSI uses Wilder smoothing, it's the same as exponential but with a longer period, there can be a noticeable discrepancy from that. "

Any conventions with larger or smaller periods, exponential or conventional smoothing - what's the point of all this? DA takes the Bull by the horns and the Bear by the paws and feet.

How do I know what you have there? What I've got I've shown. I even explained, so that it would be clearer to all who are in the tank.

[Dmitry Fedoseev]

Дмитрий:

You're trying to argue with formulas with a man who speaks the language of poetry.

:)

[Yousufkhodja Sultonov]

Дмитрий:

You are trying to argue with formulas with someone who speaks the language of poetry.

You can talk to me herehttps://www.mql5.com/ru/articles/250 and herehttps://www.mql5.com/ru/articles/1825 in the language of formulas.

[Dmitry Fedoseev]

Yousufkhodja Sultonov:
You can talk to me herehttps://www.mql5.com/ru/articles/250 and herehttps://www.mql5.com/ru/articles/1825 in the language of formulas.

And I'm talking here... and I don't draw flamastir flowers.

[Yousufkhodja Sultonov]

Dmitry Fedoseev:

And I'm talking here... and I'm not painting flamastir flowers.

This was addressed to Dimitri.

[Dmitry Fedoseev]

Refined. Enabled/disabled component smoothing. ComponentsSmooth variable. Attached to the post.

Looked at this code, especially that 'place:

bool ProcessBar(int nBarIndex, const double &farrClose[])
{
   if (nBarIndex + i_nPeriod > ArraySize(farrClose))
      return true;

   double fBullsSumm = 0.0, fBearsSumm = 0.0;
   int nBullsCnt = 0, nBearsCnt = 0;
   
   for (int i = nBarIndex + i_nPeriod - 1; i >= nBarIndex; --i)
   {
      double fPower = farrClose[i] - farrClose[i + 1];
      if (fPower > 2 * DBL_EPSILON)
      {
         fBullsSumm += fPower;
         nBullsCnt++;
      }
      if (fPower < -2 * DBL_EPSILON)
      {
         fBearsSumm -= fPower;
         nBearsCnt++;
      }
   }
   
   g_farrBullsPower[nBarIndex] = (nBullsCnt == 0)? 0.0 : 
                                  fBullsSumm / nBullsCnt / g_fPoint;
   g_farrBearsPower[nBarIndex] = (nBearsCnt == 0)? 0.0 : 
                                  fBearsSumm / nBearsCnt / g_fPoint;

   return true;
}

This is nothing more than the initial stage of calculating RSI components (with minor non-principled differences).

[Dmitry Fedoseev]

Error in my indicator (blunt). 5 min. will fix it.

[Yousufkhodja Sultonov]

Dmitry Fedoseev:

Refined. Enabled/disabled component smoothing. ComponentsSmooth variable. Attached to the post.

Looked at this code, especially that 'place:

This is nothing more than the initial stage of calculating RSI components (with minor non-principled differences).

In the given variant of DA indicator we gave a variant, which has nothing in common with RSI, the only thing in common - it's getting and using differences of adjacent price values, and that in RSI between bars, and in DA - inside of the 0th bar. Is the use of price differences tabooed by Widler as well?

[Dmitry Fedoseev]

Yousufkhodja Sultonov:
In the given variant of DA indicator we gave a variant, which has nothing in common with RSI, the only thing in common is to get and use the difference of adjacent price values, and that in RSI between bars, and in DA - inside of the 0-th bar. Is the use of price differences tabooed by Widler as well?

In RSI there is exponential smoothing, while you have a simple average.

[Dmitry Fedoseev]

Corrected (appendix).

Here is a comparison:

Yellow line - RSI components.

Red - simple smoothing with summation count (Yousufkhodja style).

Template attached.