Re: [amibroker] Re: Gravity center mustafa Belkhayate, AmiBroker Email List Archive

Hello,

The testing env for R-plugin I used is:

Windows Vista Home

AmiBroker 5.20 Real Time

ESignal Data Source

R 2.8.1 for Windows

R_Scilab_DCOM3.0-1B5..exe

rscproxy.dll (as the R Dcom documentation suggested)

RMathAFL.dll (by Patrick)

However, there is a little wrong that the plugin sometimes gives error when it seems is doing refreshing/recalculating. When I scroll on the Amibroker date axis, the error will be gone and the chart re-displayed. However, this little error is not similiar to yours.

I have no idea yet. Maybe you could compare the env, particular if we are using the same version of the Dcom server?

Rdgs,

From: Edward Pottasch <empottasch@xxxxxxxbe>
To: amibroker@xxxxxxxxxps.com
Sent: Tuesday, April 14, 2009 8:35:56 PM
Subject: Re: [amibroker] Re: Gravity center mustafa Belkhayate

thanks Steve,

that looks really simple to implement. I did not work with the RMath plugin yet. I get an error message saying: Error 47. Exception occurred during AFL formula execution at address: 77124EBC, code: C0000005

same error for the separate pieces of code.

any idea? RMath plugin is loaded..

thanks, rgds, Ed

----- Original Message -----

From: Steve Wong

To: amibroker@xxxxxxxxx ps.com

Sent: Tuesday, April 14, 2009 1:58 PM

Subject: Re: [amibroker] Re: Gravity center mustafa Belkhayate

Hello

I tried to post yesterday however my message seems got lost (I can't see it appears in the group after long hours), so I post again today, I'm sorry if this is a duplicate message:

my reply yesterday: about doing Lowess smoothing in R-plugin

Example: (please adjust the span parameter accordingly)

RMathSetArray( BarIndex( ),"x");
RMathSetArray( Close,"y" );
RMathProcedure( "y.loess <- loess(y ~ x, span=0.25, data.frame(x= x,y=y))") ;
RMathProcedure( "y.predict <- predict(y.loess, data.frame(x= x))");
Lowress = RMathGetArray( "y.predict" );

Plot(Lowress, "Lowress" , colorRed);

Polyfit: thanks Ed for the link of AFL, below is similar function in R (I hardcoded the nth order, pls adjust the polynomia order accordingly)

Example:

RMathSetArray( BarIndex( ),"x");
RMathSetArray( Close,"y" );
RMathProcedure( "y.polyfit <- lm( y ~ x + I(x^2) + I(x^3) + I(x^4)+I(x^5) + I(x^6) + I(x^7) + I(x^8))");
// RMathProcedure( "y.polyfit <- lm( y ~ x + I(x^2) + I(x^3) + I(x^4))"); // 4th order
RMathProcedure( "y.predict <- predict(y.polyfit, data.frame(x= x))");
Polyfit = RMathGetArray( "y.predict" );

Plot(Polyfit, "Polyfit" , colorRed);

regards,

SW

From: Edward Pottasch <empottasch@xxxxxxx be>
To: amibroker@xxxxxxxxx ps.com
Sent: Tuesday, April 14, 2009 6:12:01 PM
Subject: Re: [amibroker] Re: Gravity center mustafa Belkhayate

i think you are right. Fred posted some great code in the library, see: http://www.amibroke r.com/library/ formula.php? id=731

rgds, Ed

----- Original Message -----

From: Ed Hoopes

To: amibroker@xxxxxxxxx ps.com

Sent: Friday, April 10, 2009 5:04 PM

Subject: [amibroker] Re: Gravity center mustafa Belkhayate

I checked the UTube video. The upper and lower bands + the dashed lines are a least squares regression fit to a cubic equation. They are offset above and below by some measure of volatility like ATR. The yellow line in the middle is a least squares fit to a straight line (linear regression) over the same number of bars as the cubic.

I have coded a least squares fit to a quadratic in AFL - and there have been postings here also. The main problem with coding is that the x-axis time values numbers get so big that they occasionally exceed the maximum 32 bit floating point range (10^38 i think) causing the calculation to blow up. I got around this problem by copying the price/time data to separate arrays with time time array starting at zero. This stopped the code from blowing up when the numbers got too big.

With a cubic, things would be even worse since the matrix is 4x4 rather than 3x3 as in the quadratic case.

ReefBreak

--- In amibroker@xxxxxxxxx ps.com, "conrad_faber" <cfaber@xxx> wrote:
>
> The money he claims to make comes mostly from his Martingale MM....many small wins till the Tsunami strikes. Good luck, it may be worth it in the end.
>
>
> --- In amibroker@xxxxxxxxx ps.com, "Edward Pottasch" <empottasch@ > wrote:
> >
> > thanks. Yes came across oan and read somewhere about repainting. Oan claims he makes tons of money with it. But repainting is basicly looking into the future. Hard to test such a system but will study it some more,
> >
> > regards, Ed
> >
> >
> >
> > ----- Original Message -----
> > From: conrad_faber
> > To: amibroker@xxxxxxxxx ps.com
> > Sent: Friday, April 10, 2009 3:31 PM
> > Subject: [amibroker] Re: Gravity center mustafa Belkhayate
> >
> >
> >
> >
> >
> > FWIW
> >
> > The indicator that is used by oan on his FF benkyalate thread is based on the centre of gravity indicator of which there are two types. One repaints the past as new bars are formed (used by oan) and one does not. Not having used it myself I do not know which one your code relates to. Best to search FF forums.
> >
> > Regards Conrad
> >
> > --- In amibroker@xxxxxxxxx ps.com, "Edward Pottasch" <empottasch@ > wrote:
> > >
> > > thanks,
> > >
> > > yes similar to what you see on this video:
> > >
> > > http://www.youtube. com/watch? v=zD_zIkEymE0
> > >
> > > it plots the lines in real time but I can't see if it changes the lines at a later stage when information poures in. I do not "speak" mql4 language therefor I ask. But it shouldn't be too hard to translate. If somebody knows it will save me the effort.
> > >
> > > regards, Ed
> > >
> > >
> > >
> > >
> > >
> > > ----- Original Message -----
> > > From: reinsley
> > > To: amibroker@xxxxxxxxx ps.com
> > > Sent: Friday, April 10, 2009 1:56 PM
> > > Subject: Re: [amibroker] Gravity center mustafa Belkhayate
> > >
> > >
> > >
> > >
> > >
> > >
> > > I have got a CD demo running with MetaTrader 4 platform.
> > > I did not checkin to get real time quotes, but on historical datas only
> > > the last 150 bars can plot the GC.
> > > I did not count exactly, it's between 150 and 200 bars. Always the last
> > > bars and the same number of bars, you cannot move the indicator in the
> > > past, unless to remove datas.
> > >
> > > So I suppose it looks into the future. We cannot have such a smooth
> > > curve in the past and keep the same turning points.
> > >
> > > Another Belkhayate Timing indicator was supplied but it hangs the PC.
> > >
> > > However it worth to test it in real time before to write bad things
> > > about it.
> > >
> > > Best regards
> > >
> > > Edward Pottasch a écrit :
> > > >
> > > >
> > > >
> > > > looks interesting:
> > > >
> > > > http://www.tradings ystemforex. com/expert- advisors- backtesting/ 219-belkhayate- expert-advisor. html
> > > > <http://www.tradings ystemforex. com/expert- advisors- backtesting/ 219-belkhayate- expert-advisor. html>
> > > >
> > > > did not try to translate it yet. Anyone know if this code looks into the
> > > > future? Is this system of Belkhayate related to this filter:
> > > >
> > > > http://codebase. mql4.com/ 2297 <http://codebase. mql4.com/ 2297>
> > > >
> > > > ?
> > > >
> > > > thanks,
> > > >
> > > > regards, Ed
> > > >
> > > >
> > > >
> > > >
> > > > ----- Original Message -----
> > > > *From:* Michel Guibert <mailto:michelg14@ >
> > > > *To:* amibroker@xxxxxxxxx ps.com <mailto:amibroker@xxxxxxxxx ps.com>
> > > > *Sent:* Thursday, April 09, 2009 8:23 PM
> > > > *Subject:* [amibroker] Gravity center mustafa Belkhayate
> > > >
> > > > Hi
> > > >
> > > > The gravity center from Mustafa Belkhayate on Prorealtime is looking
> > > > close to the following code , anybody has an idea about converting
> > > > it in Amibroker ??
> > > > ============ ========= ========= ========= ========= ====
> > > > k=p3 // Variable p3= 65
> > > >
> > > > de48=DPO[k*2] (close)
> > > > if de48=de48[1] and de48[1]=de48[ 2] and de48[2]<>de48[ 3] then
> > > > flag=1
> > > > endif
> > > > n=(k*2)-4
> > > > p=(n/2)-1
> > > > d100=DPO[n]( close)
> > > > moy100=close- d100
> > > > co=(moy100-moy100[ 1]+(close[ p])/n)*n
> > > > if flag[1]=1 and flag[2]=0 then
> > > > hh=co[1]
> > > > endif
> > > > if flag[1]=1 then
> > > > co=hh
> > > > endif
> > > > n=p3 mod 2
> > > > p=(p3-n)/2
> > > > p3=(2*p)+1
> > > > once x=0
> > > > w=abs((p-x)/ p)
> > > > w=w*w*w
> > > > w=(1-w)
> > > > w=w*w*w
> > > > x=x+1
> > > > if barindex=p3 then
> > > > a=0
> > > > b=0
> > > > e=0
> > > > for i=1 to p3
> > > > z=barindex-i+ 1
> > > > a=a+w[z]
> > > > b=b+w[z]*(i)
> > > > e=e+(i)*(i)* w[z]
> > > > next
> > > > endif
> > > > if barindex>p3 then
> > > > c=0
> > > > d=0
> > > > for i=1 to p3
> > > > z=barindex-i+ 1
> > > > c=c+co[p3+p- i]*w[z]
> > > > d=d+co[p3+p- i]*w[z]*( i)
> > > > next
> > > > endif
> > > > alpha=(a*d-b* c)/(a*e-b* b)
> > > > beta=(c*e-b* d)/(a*e-b* b)
> > > > lowess=alpha* (p+1)+beta
> > > > if barindex < p3*2 then
> > > > lowess=undefined
> > > > endif
> > > >
> > > > zz = close - lowess
> > > > zzsigma = STD[p3](zz)
> > > > bb1 = lowess + zzsigma
> > > > bb2 = lowess +1.618* zzsigma
> > > > bb3 = lowess +2* zzsigma
> > > > bl1 = lowess - zzsigma
> > > > bl2 = lowess - 1.618* zzsigma
> > > > bl3 = lowess - 2* zzsigma
> > > >
> > > > return lowess, bb1, bb2, bb3, bl1, bl2, bl3
> > > >
> > > >
> > > > ------------ --------- --------- --------- --------- --------- -
> > > > Internet Explorer 8 makes surfing easier. Get it now!
> > > > <http://go.microsoft .com/?linkid= 9655265>
> > > >
> > > >
> > >
> >
>