Adaptive lookback indicators - page 71

William Snyder
9493
William Snyder  
camisa:
and alerts!

Camisa, added the arrows and alerts.

ssm_velocity_vhf_adaptive_-_floating_levels_alerts__arrows.mq4

kalpahlim
59
kalpahlim  
mladen:
Yes, but there are also versions that are using ATR too. Implicitly those are adaptive versions (since the ATR will change from bar to bar), but the ATR itself is also not suitable for adapting

Hi,

I am wondering if a relatively simple indicator such as the double-smoothed Jurik histo could be made adaptive with ATR? Obviously the ATR itself will need to be averaged a little with periods.I am thinking the higher the ATR, the lower the periods for the Jurik histo need to be, so it would be an inverse relationship, designed to try to take advantage of rapid price movements with higher ATR, and then lengthen the indicator's periods when the market is flat. I presume this is the sort of thing that 'adaptive' does anyway???

Anyway would like your thoughts on this mladen and mrtools, if possible.

Thanks in advance.

S

Mladen Rakic
160767
Mladen Rakic  
seerkat:

Hi,

I am wondering if a relatively simple indicator such as the double-smoothed Jurik histo could be made adaptive with ATR? Obviously the ATR itself will need to be averaged a little with periods.I am thinking the higher the ATR, the lower the periods for the Jurik histo need to be, so it would be an inverse relationship, designed to try to take advantage of rapid price movements with higher ATR, and then lengthen the indicator's periods when the market is flat. I presume this is the sort of thing that 'adaptive' does anyway???

Anyway would like your thoughts on this mladen and mrtools, if possible.

Thanks in advance.

S

One problem is the following : we, the men, can see when something is high or low. How does the code for that ATR "knows" if the ATR is high or low? Without some normalization, that can not be known, and any normalization deforms the values

myname
3231
myname  
mladen:
Yes, but there are also versions that are using ATR too. Implicitly those are adaptive versions (since the ATR will change from bar to bar), but the ATR itself is also not suitable for adapting

Yes, I see. What if we sued the standard deviations for that?

SilverBB
9
SilverBB  
mladen:
One possible issue in the t3 - vhf adaptive - removed : t3_vhf_adaptive_1_3.ex4

Hi mladen,

Could we have a Histo version of this please?

Thanks

kalpahlim
59
kalpahlim  
mladen:
One problem is the following : we, the men, can see when something is high or low. How does the code for that ATR "knows" if the ATR is high or low? Without some normalization, that can not be known, and any normalization deforms the values

I presume this has been tried; but for example with this indicator on M5, could we have the periods of the jurik vary from about 48 when the ATR(5) is 5 pips to about 8 when the ATR(5) is 15 pips... I know it has to be arbitrary and inputted, unless you do a proportion of a long-term average of ATR, eg 100-500, instead. I realize this might sound simplistic... but I realize the problem with indicators in general is that the 'wavelength' of the market varies so much and indicators can work well when the wavelength is right, but can be precisely the wrong wavelength when the market changes... I guess you know all this though.

Mladen Rakic
160767
Mladen Rakic  
seerkat:
I presume this has been tried; but for example with this indicator on M5, could we have the periods of the jurik vary from about 48 when the ATR(5) is 5 pips to about 8 when the ATR(5) is 15 pips... I know it has to be arbitrary and inputted, unless you do a proportion of a long-term average of ATR, eg 100-500, instead. I realize this might sound simplistic... but I realize the problem with indicators in general is that the 'wavelength' of the market varies so much and indicators can work well when the wavelength is right, but can be precisely the wrong wavelength when the market changes... I guess you know all this though.

That is too much of fixed values (not just that it would work only for one symbol, but it would work only on one time frame of that symbol, since the ATR would vary very much for other time frames of that same symbol). We need to have some kind of mathematical dependency that would be then used for the purpose of adapting

kalpahlim
59
kalpahlim  
mladen:
That is too much of fixed values (not just that it would work only for one symbol, but it would work only on one time frame of that symbol, since the ATR would vary very much for other time frames of that same symbol). We need to have some kind of mathematical dependency that would be then used for the purpose of adapting

I wasn't implying fixed values, rather a variable that ranged in a linear fashion depending on the ATR value. Perhaps a curve rather than a straight line would be a better fit; but for now:

eg Indicator Period = 200 / ATR(5) - using M5

What do you think?

Mladen Rakic
160767
Mladen Rakic  
seerkat:

I wasn't implying fixed values, rather a variable that ranged in a linear fashion depending on the ATR value. Perhaps a curve rather than a straight line would be a better fit; but for now:

eg Indicator Period = 200 / ATR(5) - using M5

What do you think?

seerkat

Will check if any of the normalizing methods can transform ATR into a range that then can be used for adapting, but can not guarantee anything

Mladen Rakic
160767
Mladen Rakic  
mladen:
Here is a version in which not just ADXm is adaptive but the floating zero level is adaptive too : adxm_vhf_adaptive_floating_zero.mq4

________________

So, this is a kind of an extreme adaptive indicator. Adapting the zero level makes the level a kind of a an indicator too

Extended version with two more levels added to it : adxm_vhf_adaptive_floating_zero_1_2.mq4