Adaptive lookback indicators - page 75

New comment
 
mladen:
Just an example that not everything is peaches when experimenting with adapting

Here is an example of Sperman rank correlation made adaptive (the rank itself is made adaptive), and, as you can see(lower is the "adaptive"), it is far from what we should qualify as successful (and improvement) : dynamic_zone_adaptive_spearman_rank_correlation_2_1.mq4

_______________

Obviously the direction I took in making Spearman adaptive was wrong and it needs to be directed in other direction. Back to the drawing board

Dearest MLADEN,

thanks much for doubling your work and consequently you find a fine solution,great ...... is this a new subject and method of adapting,other than you were using/applying recently (vhf).

regards

 
mntiwana:

Dearest MLADEN,

thanks much for doubling your work and consequently you find a fine solution,great ...... is this a new subject and method of adapting,other than you were using/applying recently (vhf).

regards

mntiwana

Actually, after doing some more tests, I do not think that Spearman rank correlation can be adapted in an acceptable way. Adapting (using any adapting type) of the rank gives that kind of result from the above example. And adapting prices, produces too much lag in the case of Spearman - so, all in all, I think it will stay as is

 
mladen:

mntiwana

Actually, after doing some more tests, I do not think that Spearman rank correlation can be adapted in an acceptable way. Adapting (using any adapting type) of the rank gives that kind of result from the above example. And adapting prices, produces too much lag in the case of Spearman - so, all in all, I think it will stay as is

Why do you think it can not be adaptive?

 
apprentice coder:
Why do you think it can not be adaptive?

Simply because of the way it is calculated. The rank "inherits" to some extent and is calculated using integer counters - can not be made different. The "inheritance" is happening because of the sorting prior to ranking and that adds to that jumpy look that is already made by integer counters mode. As long as it must be sorted as a part of calculation, nothing can be done to avoid the described

 
mladen:
Just an example that not everything is peaches when experimenting with adapting

Here is an example of Sperman rank correlation made adaptive (the rank itself is made adaptive), and, as you can see(lower is the "adaptive"), it is far from what we should qualify as successful (and improvement) : dynamic_zone_adaptive_spearman_rank_correlation_2_1.mq4

_______________

Obviously the direction I took in making Spearman adaptive was wrong and it needs to be directed in other direction. Back to the drawing board

Pity that it does not work for this indicator

 
mladen:
Simply because of the way it is calculated. The rank "inherits" to some extent and is calculated using integer counters - can not be made different. The "inheritance" is happening because of the sorting prior to ranking and that adds to that jumpy look that is already made by integer counters mode. As long as it must be sorted as a part of calculation, nothing can be done to avoid the described

Ah OK. Makes sense now

 

Dear Mladen,

could you please make the attached indicator adaptive,

possibly in the MQ4 format and without adding new moving averages?

The indicator was coded by you a few years ago.

 
wojtek.paul:
Dear Mladen,

could you please make the attached indicator adaptive,

possibly in the MQ4 format and without adding new moving averages?

The indicator was coded by you a few years ago.

Wojtek.paul

as for compatible with new mt4 i hope it will be work now, errors removed (but be care still few warning exist),and as for adapting with some thing else that moderators can better guide you.

regards

supertrend_averages_ampamp_alerts.mq4

 

Thank you, dear Mntiwana. Yes, it would be interesting to test the adaptive version, but it's up to our moderators.

 
wojtek.paul:
Thank you, dear Mntiwana. Yes, it would be interesting to test the adaptive version, but it's up to our moderators.

wojtek.paul

It will be made with the new set of averages (double smoothed and adaptive)

1...686970717273747576777879808182...128
New comment