|
PureBytes Links
Trading Reference Links
|
--- In amibroker@xxxxxxxxxxxxxxx, Scott Cong <scong@xxxx> wrote:
> DIMITRIS TSOKAKIS wrote:
>
> >Here is the 2nd, 3rd and 4th degree Q_averages.
> >The 4th degree is more complicated but gives some interesting
> >alternatives.
> >a. The Ti4 is still smooth and much faster than the 2nd and 3rd
> >degree lines.
> >b. The Ti4f may act as a magnifier, peaks are higher and troughs
are
> >lower than the actual price.
> >After this level, a significant distorsion becomes more and more
> >apparent .
> >
> >Plot(C,Date()+", C",colorBlack,128);
> >p=20;s=0.7;
> >e1=EMA(C,p);
> >e2=EMA(e1,p);
> >e3=EMA(e2,p);
> >e4=EMA(e3,p);
> >e5=EMA(e4,p);
> >e6=EMA(e5,p);
> >//2nd degree coefficients
> >d1=s^2;
> >d2=-2*s*(s+1);
> >d3=(1+s)^2;
> >Ti2=d1*e6+d2*e5+d3*e4;Plot(Ti2,"\nTi2",colorRed,1);
> >Ti2f=d1*e5+d2*e4+d3*e3;
> >Plot(Ti2f,"\nTi2f",colorPink,1);
> >Ti2ff=d1*e4+d2*e3+d3*e2;
> >Plot(Ti2ff,"\nTi2ff",colorOrange,1);
> >//3rd degree coefficients
> >C1=-s^3;
> >C2=3*s^2*(1+s);
> >C3=-3*s*(s+1)^2;
> >C4=(1+s)^3;
> >Ti3=c1*e6+c2*e5+c3*e4+c4*e3;
> >Plot(Ti3,"\nTi3",colorDarkGreen,8);
> >Ti3f=c1*e5+c2*e4+c3*e3+c4*e2;
> >Plot(Ti3f,"\nTi3f",colorBrightGreen,1);
> >//4th degree coefficients
> >g1=s^4;
> >g2=-4*(s+1)*s^3;
> >g3=6*(s+1)^2*s^2;
> >g4=-4*(s+1)^3*s;
> >g5=(s+1)^4;
> >Ti4=g1*e6+g2*e5+g3*e4+g4*e3+g5*e2;
> >Plot(Ti4,"\nTi4",colorDarkGrey,8);
> >Ti4f=g1*e5+g2*e4+g3*e3+g4*e2+g5*e1;
> >Plot(Ti4f,"\nTi4f",colorLightGrey,8);
> >
> >
> >
> >
> Just a thought.
>
> In automatic control field, one constructs PID (proportiaonal,
integral
> and derivative) controllers to regulate the target based on the
> derivation between the desired output and the real output. The
integral
> component is responsible for the history, thus a lagging factor
(moving
> average is just the one), the derivative part (Rate of Change) is a
> prediction for the future, thus a leading factor causing more
wipsaws;
> whileas the proportional component focuses on the current status.
>
> Combining them together offers a compromise of responsiveness and
> smoothness. The ideal one is to have minimum lag, minimum overshoot
and
> undershoot. This is just the requirement for ideal moving averages.
Scott,
This "Combining" is the magic word of course...
In daily data the shortest dt is equal to 1 bar. This limitation
keeps the actual derivative dQ/dt far from the theoretical one.
The closest approximation is perhaps the Relative Slope
http://www.amibroker.com/library/detail.php?id=43
As for the integral, there was an interesting discussion two years
ago, as you may see in
http://finance.groups.yahoo.com/group/amibroker/message/28159
http://finance.groups.yahoo.com/group/amibroker/message/28182
http://finance.groups.yahoo.com/group/amibroker/message/28220
http://finance.groups.yahoo.com/group/amibroker/message/28224
The 1 bar limitation will also introduce an error in integration
procedures.
See, for example, the
x=Cum(1)-1;
y=x^2;Plot(Cum(y),"calculated Integ",1,1);
integ=x^3/3;Plot(integ,"Integ",4,1);
Plot(100*(Cum(y)-integ)/integ,"error",2,styleHistogram+styleOwnScale);
We need 150 bars for error<1% but, even after 1000 bars the error is
still 0.15% [for the example function]
Dimitris
------------------------ Yahoo! Groups Sponsor --------------------~-->
Make a clean sweep of pop-up ads. Yahoo! Companion Toolbar.
Now with Pop-Up Blocker. Get it for free!
http://us.click.yahoo.com/L5YrjA/eSIIAA/yQLSAA/GHeqlB/TM
--------------------------------------------------------------------~->
Check AmiBroker web page at:
http://www.amibroker.com/
Check group FAQ at: http://groups.yahoo.com/group/amibroker/files/groupfaq.html
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/amibroker/
<*> To unsubscribe from this group, send an email to:
amibroker-unsubscribe@xxxxxxxxxxxxxxx
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
|