Trying to find the cause of thinness... - page 3

New comment
 
Lorentzos Roussos #:

😊

Dear Lorentzos, by now the indicator returns a value that changes depending on the timeframe and average Period. Would it be possible to make it as a ratio that oscillates always between 2 values like 0-100, independently from the timeframe or average period used? I mean, a normalization...
 
Gtrade #: Dear Lorentzos, by now the indicator returns a value that changes depending on the timeframe and average Period. Would it be possible to make it as a ratio that oscillates always between 2 values like 0-100, independently from the timeframe or average period used? I mean, a normalization...

If you try to "normalise" it, then you will lose the original aim of your question in your first post.

I suggest you take some time to research the maths and code presented so far, and consider their meaning, specially in regards to the data being analysed.

You can't just go making changes without regard for the underlying meaning and the data being used for the calculations. Whatever changes you want to make, should make "sense" statistically.

 
Fernando Carreiro #:

If you try to "normalise" it, then you will lose the original aim of your question in your first post.

I suggest you take some time to research the maths and code presented so far, and consider their meaning, specially in regards to the data being analysed.

You can't just go making changes without regard for the underlying meaning and the data being used for the calculations. Whatever changes you want to make, should make "sense" statistically.

That's the reason why I made a question, beginning with "Would it be" and ending with a question mark. Not all are mathematicians or coders, since it is enough difficult and takes enough time to be a good trader...instead you, should ask yourself the question: "why when I say something I'm not able to avoid to do it in a rude manner?"... I was sure that my previous words in this thread had implicitly but clearly meant that your words in my threads are not welcome.

 

Come on guys don't fight . At some point the sun will explode and swallow the solar system , our job is to have as much fun until then  😊

Gtrade #:
Dear Lorentzos, by now the indicator returns a value that changes depending on the timeframe and average Period. Would it be possible to make it as a ratio that oscillates always between 2 values like 0-100, independently from the timeframe or average period used? I mean, a normalization...

I set the plots 0.0 to 1.0 in my first attempt but i did not deem that to be significant ,there seems to be a "factor" that holds the oscillation in a region for some duration of time , perhaps that is the initial indication you were looking for ? 

 
Lorentzos Roussos #: I set the plots 0.0 to 1.0 in my first attempt but i did not deem that to be significant ,there seems to be a "factor" that holds the oscillation in a region for some duration of time , perhaps that is the initial indication you were looking for ? 

Given that you are using the ratio [abs(close-to-close)] / [true-range], then it is already normalised for the values from 0 to 1, given that [abs(close-to-close)] <= [true-range].

However, since this "normalisation" can only seen from a zoomed-out "big picture" sense of things, it may seem that it is not normalised at a zoomed-in view.

Trying to "normalise" it at the zoomed-in level, will cause the ratio to lose its value, and no longer be able indicate a true measure of inverse deviation/volatility, or "thinness" or what-ever you wish to call it.

However, given that my input has no value or even wanted by for the OP, I am sure that if you can come up with a method for doing "local normalisation", then the OP will be very happy with it.
 
Fernando Carreiro #:

Given that you are using the ratio [abs(close-to-close)] / [true-range], then it is already normalised for the values from 0 to 1, given that [abs(close-to-close)] <= [true-range].

However, since this "normalisation" can only seen from a zoomed-out "big picture" sense of things, it may seem that it is not normalised at a zoomed-in view.

Trying to "normalise" it at the zoomed-in level, will cause the ratio to lose its value, and no longer be able indicate a true measure of inverse deviation/volatility, or "thinness" or what-ever you wish to call it.

However, given that my input has no value or even wanted by for the OP, I am sure that if you can come up with a method for doing "local normalisation", then the OP will be very happy with it.

yeah this . So when i zoomed out i saw it was at the low end , this is the indication essentially , right ? An asset being at the low end indicates low thinness if compared to another one .

 
@Lorentzos Roussos #:yeah this . So when i zoomed out i saw it was at the low end , this is the indication essentially , right ? An asset being at the low end indicates low thinness if compared to another one .

The closer it is to zero (at the extreme end), means one of two possible things:

  1. price is maintaining a stable mean value (going sideways, almost no trending).
  2. volatility is extremely high, compared to the overall price mean (even if it is trending)
     
    Lorentzos Roussos #: yeah this . So when i zoomed out i saw it was at the low end , this is the indication essentially , right ? An asset being at the low end indicates low thinness if compared to another one .

    The closer it is to one (at the extreme end), means that price is more pure and/or volatility is probably extremely low.

       
      Fernando Carreiro #:

      The closer it is to zero (at the extreme end), means one of two possible things:

      1. price is maintaining a stable mean value (going sideways, almost no trending).
      2. volatility is extremely high, compared to the overall price mean (even if it is trending)

        yeah , that means i should remove the 2 other lines in your calc mode as they don't indicate anything .

        But , in the first calculation mode it appears the blue line indicates the "purity" of the trend and the red line the "willingness" to counter it.

        But (2) , the 2 lines have a flaw that when their condition is not met the previous value is carried . 
         
        Lorentzos Roussos #: yeah , that means i should remove the 2 other lines in your calc mode as they don't indicate anything. But , in the first calculation mode it appears the blue line indicates the "purity" of the trend and the red line the "willingness" to counter it.But (2) , the 2 lines have a flaw that when their condition is not met the previous value is carried . 

        I don't really want to get involved any further and then be accused of being rude for my contribution. So, I will leave further analysis to you and the OP.

        1234
        New comment