Digital ACSTrend - page 42

 
techmac:
Why does it look like an inverted momentum?

Lyapunov Exponents measure the change in divergence of the track of a given motion/particle.

Under Lyapunov formula, there is simplistically (ignoring natural logarithmn etc), change in price divided by original price.

The general formula for momentum is Price Y divided by Price X in different periods.

& since we are using in an "online" manner for "real-time usages".... the original price will keep updating with time too... resulting in change in price increases highly likely being correlated with Compared Price...

So that's what makes them all too similar, with primary difference being the scale....

For practical purposes, the formulas are almost the same essentially.

A more simplified explaination of Lyapunov:

4.3 Lyapunov Exponent

Wintersky

Cheers

 
wintersky111:
Lyapunov Exponents measure the change in divergence of the track of a given motion/particle.

Under Lyapunov formula, there is simplistically (ignoring natural logarithmn etc), change in price divided by original price.

The general formula for momentum is Price Y divided by Price X in different periods.

& since we are using in an "online" manner for "real-time usages".... the original price will keep updating with time too... resulting in change in price increases highly likely being correlated with Compared Price...

So that's what makes them all too similar, with primary difference being the scale....

For practical purposes, the formulas are almost the same essentially.

A more simplified explaination of Lyapunov:

4.3 Lyapunov Exponent

Wintersky

Cheers

Thanks for the additional information

What I was wondering is why a 1 period inverted momentum :

LyanupovBuffer= MathLog (MathAbs(iMA(NULL,0,LyapunovTime,0,0,0,ii+1)/iMA(NULL,0,LyapunovTime,0,0,0,ii)));

is named Lyapunov exponent (MathAbs() is not needed in that line since prices are never negative and MathLog() is just changing the scale of the resulting values).

Anyway, thanks again

 
techmac:
Thanks for the additional information

What I was wondering is why a 1 period inverted momentum :

LyanupovBuffer= MathLog (MathAbs(iMA(NULL,0,LyapunovTime,0,0,0,ii+1)/iMA(NULL,0,LyapunovTime,0,0,0,ii)));

is named Lyapunov exponent (MathAbs() is not needed in that line since prices are never negative and MathLog() is just changing the scale of the resulting values).

Anyway, thanks again

Isn't the time parameter at a wrong place in that calculation?

 
nbtrading:
Isn't the time parameter at a wrong place in that calculation?

Guys

Forget about that indicator - that was probably an experiment that went wrong

Use information provided by nevar and wintersky111 to obtain accurate information of what Lyapunov exponent is (and what can it be used for)

 
mladen:
Guys

Forget about that indicator - that was probably an experiment that went wrong

Use information provided by nevar and wintersky111 to obtain accurate information of what Lyapunov exponent is (and what can it be used for)

OK

But that indicator is misleading - I am not some coder but sometimes some code makes me wonder what was the intention - like parameters that are never used (there are 2 parameters only one is used for something meaningful) and declaring 3 buffers with only 1 used

__________

Anyway, thanks to all that pointed to correct information - chapter closed for me now

 
nbtrading:
OK

But that indicator is misleading - I am not some coder but sometimes some code makes me wonder what was the intention - like parameters that are never used (there are 2 parameters only one is used for something meaningful) and declaring 3 buffers with only 1 used

__________

Anyway, thanks to all that pointed to correct information - chapter closed for me now

Good

As it is clear now there are correctly coded versions - so, all should be OK. We all were beginers at some time when we had no idea at all what are we doing at what we are doing - so, leave that as is and use the stuff that is done as it should be. And everybody will be happy ever after

 
mladen:
Good As it is clear now there are correctly coded versions - so, all should be OK. We all were beginers at some time when we had no idea at all what are we doing at what we are doing - so, leave that as is and use the stuff that is done as it should be. And everybody will be happy ever after

The magic of internet - we can find anything

 
mladen:
Good As it is clear now there are correctly coded versions - so, all should be OK. We all were beginers at some time when we had no idea at all what are we doing at what we are doing - so, leave that as is and use the stuff that is done as it should be. And everybody will be happy ever after

Sorry for getting back to this : checked some more stuff coded by the same guy, and even though I am an amateur coder, some of the code is even worse. Check this : https://www.mql5.com/en/forum/180729/page19 . That thing is doing anything but what it tells to be doing. Now I am avoiding anything coded by him

 

Lyapunov HP oscillator

lyapunov_hp.mq4

Files:
 
tampa:
Re: Digital ACSTrend

Lyapunov HP oscillator

lyapunov_hp.mq4

I know that it repaints, but can it be made a bit more efficient?