Re: [amibroker] Re: Ehlers Dominant Cycle

[Rakesh Sahgal]

Is the book by Brian Millard , which is supposed to be based on Hurst's
work,  any help in  understanding Hurst's work  - a sort
of  Hurst for dummies kind of thing?

Rakesh On 10/7/06, Fred <[EMAIL PROTECTED]> wrote:
There are other books that would no doubt provide some clues but Idon't think Hurst provides any additional clues in his book to what'sin the appendecies which are I think back up information /methodologies for the book.
--- In amibroker@yahoogroups.com, "Steve Dugas" <[EMAIL PROTECTED]> wrote:>> Hi Fred - Well, I just opened the book for the first time to look
at App 6> and I'm afraid  its over my head. Maybe I will have a better graspafter> reading the book, but unfortunately I am not an engineer oranything, and my> first thought is that if it is giving *you* some trouble, then *my*
chances> are probably slim to none.   8 - )   But anyway, I will do somereading over> the weekend, and see if I can learn anything...>> BTW, I noticed that our two threads seemed to be related. It would
seem that> I am starting down a road that you have some experience in so Iwould value> your opinion. As far as timing is concerned, do you think thatcycles, ala> Ehlers, Hurst, etc is the preferred way to go? In the future, after
I have> come up to speed, perhaps we could bounce some ideas around, etc...>> Steve>> ----- Original Message -----> From: "Fred" <[EMAIL PROTECTED]>> To: <
amibroker@yahoogroups.com>> Sent: Friday, October 06, 2006 2:26 PM> Subject: [amibroker] Re: Ehlers Dominant Cycle>>> > Steve,> >
> > Are your math skills strong enough to implement what's inAppendix 6> > of Hurst's book ?> >> > --- In amibroker@yahoogroups.com
, "Steve Dugas" <sjdugas@> wrote:> >>> >> OK, thanks Andy!  I will do something else for a few days andwait> > for Cybernetics to arrive ( I also got the MESA book, but had to
> > wait while they backordered both. )  I have had Hurst's book on my> > shelf for a couple of months now, waiting patiently for a little> > attention - I will look that over in the meantime. Thanks very
much> > and good luck with your move - maybe we can pick this up again a> > little way down the road...> >>> >> Steve> >>> >> ----- Original Message -----
> >>   From: Andy Davidson> >>   To: amibroker@yahoogroups.com> >>   Sent: Friday, October 06, 2006 4:27 AM> >>   Subject: Re: [amibroker] Ehlers Dominant Cycle
> >>> >>> >>   Steve,> >>> >>   I think Ehler believes that the Cyber Cycle method isbetter...I> > remember that being my impression when I read the books. Certainly
> > both approaches use the Hilbert Transform but I think thedifference> > lies in the method of extracting the Quadrature and In-Phase> > components. Difficult to say exactly without the book to hand to
> > double check what I'm saying!> >>> >>   As for the differences between approaches, try playing around> > with the alpha parameter (which determines the smoothing and> > therefore low-end cut-off level for the measured cycles). You will
> > probably see some very large shifts in the period for even a 0.01> > change in alpha. This tells me something.> >>> >>   I do use Ehler's cycle measurements in some of my indicators as
> > an adaptive input to the period function. He talks about this> > approach in the Cybernetic book. The reason I do this is becausemy> > feeling is that something which at least attempts to measure a
cycle> > period is probably better than than some arbitrary optimisednumber.> > I say 'probably' as I can't prove this to be the case - it's justa> > matter of what makes more logical sense to me.
> >>> >>   To my mind though there's one pretty big hole in Ehler's> > approach. Basically he states that although there are multiple> > cycles at play in the market at any one time there is only one
that> > is "dominant" and thus tradeable. So you are only taking into> > account and trying to measure one specific cycle at any one time.> > One way he achieves this is through limiting the parameters to
> > specific wavelength ranges. Therefore you might be assuming thatyou> > are (a) going to ignore cycles less than 6-bars in period as noise> > and (b) going to ignore cycles of more than, say, 60 bars as too
> > long to trade (or too prone to margins of error) and thenmeasuring> > *the* cycle in that range. Well what if there is a 20-bar and a45-> > bar cycle at play at the same time and the noise sometimes
exceeds 6-> > bars? My experience is that Ehler's method is not good at coping> > with this. I'm no engineer and so my theorising might be flawed,but> > my experience of engineering tells me that you get to work with
much> > bigger margins of error than we can tolerate as traders. For> > example, you might be able to use Hilbert Transforms in electronic> > engineering to extract a person's voice from a 'noisy'
> > waveform...but I would say that there is still too much noise inthe> > output signal to transfer the analogy to the trading world. We can> > make sense of the extracted voice but this is because our human
> > brains are good at that sort of thing, not because the modified> > signal is particularly clear in a real sense. Our trading capitalis> > not so good at dealing with the remnant noise!!
> >>> >>   As far as measuring cycles goes, I have had much more success> > and have much more confidence in the approach outlined by Hurst in> > his "Profit Magic of Stock Transaction Timing". An old book which
> > uses centred MAs and assumes there are many cycles to be> > measured...centred MAs are much more low-tech than Hilbert> > Transforms I know, but there's beauty in the simplicity if you can
> > get away from the idea that your indicators must tell you> > the "answer" right up to the right edge of the chart. Of course,the> > fact that centred MAs *don't* go the right-edge makes backtesting
> > very difficult. Well, there's flaws with everything of course...> >>> >>   Anyway, let me know how you get on and I'll help more if I can.> > I'm in the process of moving house at the moment so can't do much
> > more until I get set-up again. Good luck.> >>> >>   Andy> >>> >>> >>   Steve Dugas wrote:> >>     Hi Andy,> >>> >>     In Rocket Science, Ehlers shows 3 methods to compute the
cycle> > period. Then he tests them against each other and determines that> > the Homodyne Discriminator method had the the best characteristics> > of the 3, so he uses that code as a basis for the indicators that
> > follow. The 3 yeilded fairly similar results, so when my graph> > looked wrong, I figured the code from the library should give me a> > rough idea of what mine should look like even though it is created
> > through a different technique. I still think there is something> > wrong with mine, it should look much more like the other one.Maybe> > I will be able to just substitute your code for mine and build
from> > there. Is the one in Cybernetics supposed to be better? I ordered> > that book too, but it got delayed and I just found out that it> > shipped today. Thanks for the code!> >>
> >>     Steve> >>       ----- Original Message -----> >>       From: Andy Davidson> >>       To: amibroker@yahoogroups.com
> >>       Sent: Thursday, October 05, 2006 1:46 PM> >>       Subject: Re: [amibroker] Ehlers Dominant Cycle> >>> >>> >>       Steve,> >>       I think you're getting confused between Ehler's two books.
> > As far as I recall (books not to hand right now) he makes the> > confusion easy as there is a "Dominant Cycle" indicator in> > both 'Rocket Science' and 'Cybernetic', which are based on
different> > methods. Looks to me like the top one on your plot is the formerand> > the bottom is the latter. I personally have used the latter...the> > code is copied below, which is probably nearly identical to the
> > posted AFL library version as I used that as a starting pointwhen I> > worked through it myself.> >>       Can't help you with the 'Rocket Science' version I'mafraid.> >>       Andy
> >>> >>> >>       // Ehler's Dominant Cycle Period> >>       // Cybernetic Analysis for Stocks and Futures> >>       // Chapter 9, p. 107. Code on p. 111.
> >>> >>       //Global Parameters> >>       X = Param("MP[1] Close[2]",1,1,2,1);> >>       Z1 = IIf(X==1, (H+L)/2 , C);> >>       Z2 = Param("Alpha", .07, .01, 1, .01);
> >>> >>       function CyclePeriod(price, alpha)> >>       {> >>        instperiod = deltaphase = cycle = period = 0;> >>        Cycle = ( price[2] - 2*price[1] + price
> > [0] )/4;   //initialise arrays> >>        smooth = ( price + 2*Ref(price,-1) + 2*Ref(price,-2) + Ref> > (price,-3) )/6;> >>> >>        for (i=6 ; i<BarCount ; i++)
> >>         {>
>>          Cycle[i]
= (1-alpha/2)^2 * ( smooth[i] - 2*smooth[i-1] +> > smooth[i-2] ) +>
>>                    
2*(1-alpha)*Cycle[i-1] - (1-alpha)^2*Cycle[i-> > 2];> >>> >>          Q1[i] = (.0962*cycle[i] + .5769*cycle[i-2] -.5769*cycle[i-> > 4] - .0962*cycle[i-6])*(.5 + .08*InstPeriod[i-1]);
> >>          I1[i] = cycle[i-3];> >>> >>          if(Q1[i] != 0 AND Q1[i-1] != 0)> >>          DeltaPhase[i] = (I1[i]/Q1[i] - I1[i-1]/Q1[i-1])/(1 + I1[i]> > *I1[i-1]/(Q1[i]*Q1[i-1]));
>
>>          //limit
Delta Phase High/Low (0.09rads = 69bars, 1.1rads> > = 6bars...per page 117)> >>          if(DeltaPhase[i] < 0.09)> >>           DeltaPhase[i] = 0.09;> >>          if(DeltaPhase[i] > 
1.1)> >>           DeltaPhase[i] = 1.1;> >>>
>>          //---Begin
median calculation (placed inline for speed).>
>>          //Hardcoded
as length=5 as higher values would be out of> > range due to start-up period in main loop> >>          for(k=4; k>=0; k--)> >>           { temparray[k] = DeltaPhase[i-k]; } //create new array
> > with last 5 values of DeltaPhase> >>          temp = 0;>
>>          for(k=4;
k>0; k--)  //this series of loops re-organises> > temparray into ascending order> >>          { for (j=4; j>0; j--)>
>>            {
if (temparray[j-1] > temparray[j]) //swap values in> > array if previous value is greater> >>              { temp = temparray[j-1];>
>>                temparray[j-1]
= temparray[j];>
>>                temparray[j]
= temp;> >>          }}}> >>          MedianDelta[i] = temparray[2]; //returns the middle> > (third) element of temparray> >>          //---End median calculation> >>
> >>          DC[i] = Nz( 6.28318 / MedianDelta[i] + .5, 15 );> >>> >>          InstPeriod[i] = .33*DC[i] + .67*InstPeriod[i-1];> >>          Period[i] = .15*InstPeriod[i] + .85*Period[i-1];
> >>          }> >>        for (i=0; i<7; i++)> >>         { Period[i] = 1; }> >>        return Period;> >>       }> >>> >>       Plot( CyclePeriod(Z1,Z2) , "CyberCycle", colorRed );
> >>> >>> >>> >>       Steve Dugas wrote:> >>         Hi All,> >>> >>         I wonder if anyone has ever tried to code Ehlers Dominant
> > Cycle - the one based on the Homodyne Discriminator, pp. 68-69 in> > Rocket Science. I have never used TradeStation and this is myfirst> > shot at translating EasyLanguage. As far as I can see the code
looks> > OK to me but what do I know? Anyway, the graph it produces (middle> > one ) looks pretty bad. For comparison, I plotted the DominantCycle> > code from the AFL library on the bottom ( but I believe this uses
a> > different method ).  I would like to go on and code the rest ofthe> > indicators in the book but many are built on this so I need to get> > this right first. Any thoughts or working code would be greatly
> > appreciated. I have enclosed my code below.Thank you!> >>> >>         Steve> >>> >>         // Dominant Cycle> >>> >>         SetBarsRequired( 10000, 10000 );
> >>> >>         // USER DEFINED PARAMS> >>> >>         Price = ( High + Low ) / 2;> >>> >>         // FORMULA> >>> >>         // initialize variables
> >>> >>         Smooth = Detrender = I1 = Q1 = jI = jQ = I2 = Q2 = Re =Im> > = Period = SmoothPeriod = 0;> >>> >>         // calculate dominant cycle period
> >>> >>         for ( i = 6; i < BarCount; i++ )> >>> >>         {> >>> >>         // smooth price data with 4-bar WMA> >>
> >>         Smooth[i] = ( 4 * Price[i] + 3 * Price[i-1] + 2 * Price[i-> > 2] + Price[i-3] ) / 10;> >>> >>         // compute amplitude correction> >>> >>         AmpCorr[i] = 
0.075 * Period[i-1] + 0.54;> >>> >>         // compute detrended price data and Quadrature component> > with 7-bar Hilbert Transform> >>> >>         Detrender[i] = ( 
0.0962 * Smooth[i] + 0.5769 * Smooth[i-> > 2] - 0.5769 * Smooth[i-4] - 0.0962 * Smooth[i-6] ) * AmpCorr[i];> >>> >>         Q1[i] = ( 0.0962 * Detrender[i] + 0.5769 * Detrender[i-2] -
> > 0.5769 * Detrender[i-4] - 0.0962 * Detrender[i-6] ) * AmpCorr[i];> >>> >>         // compute InPhase component by referencing center bar of> > Hilbert Transformer ( 3 bars ago )
> >>> >>         I1[i] = Detrender[i-3];> >>> >>         // advance the phase of I1 and Q1 by 90 degrees with 7-bar> > Hilbert Transform> >>
> >>         jI[i] = ( 0.0962 * I1[i] + 0.5769 * I1[i-2] - 0.5769 * I1> > [i-4] - 0.0962 * I1[i-6] ) * AmpCorr[i];> >>> >>         jQ[i] = ( 0.0962 * Q1[i] + 0.5769 * Q1[i-2] - 
0.5769 * Q1> > [i-4] - 0.0962 * Q1[i-6] ) * AmpCorr[i];> >>> >>         // perform Phasor addition for 3-bar averaging> >>> >>         I2[i] = I1[i] - jQ[i];
> >>> >>         Q2[i] = Q1[i] + jI[i];> >>> >>         // smooth the I and Q components> >>> >>         I2[i] = 0.2 * I2[i] + 0.8 * I2[i-1];> >>
> >>         Q2[i] = 0.2 * Q2[i] + 0.8 * Q2[i-1];> >>> >>         // apply the Homodyne Discriminator> >>> >>         Re[i] = I2[i] * I2[i-1] + Q2[i] * Q2[i-1];
> >>> >>         Im[i] = I2[i] * Q2[i-1] - Q2[i] * I2[i-1];> >>> >>         // smooth the Re and Im components> >>> >>         Re[i] = 0.2 * Re[i] + 
0.8 * Re[i-1];> >>> >>         Im[i] = 0.2 * Im[i] + 0.8 * Im[i-1];> >>> >>         // compute Dominant Cycle period> >>> >>         if ( Im[i] != 0 AND Re[i] != 0 )
> >>> >>         Period[i] = 360 / atan( Im[i] / Re[i] );> >>> >>         // limit ROC of the cycle period to +/- 50% of previous> > cycle period> >>
> >>         if ( Period[i] > 1.5 * Period[i-1] )> >>> >>         Period[i] = 1.5 * Period[i-1];> >>> >>         if ( Period[i] < 0.67 * Period[i-1] )
> >>> >>         Period[i] = 0.67 * Period[i-1];> >>> >>         // limit the cycle period to be > 6 or < 50> >>> >>         if ( Period[i] < 6 )
> >>> >>         Period[i] = 6;> >>> >>         if ( Period[i] > 50 )> >>> >>         Period[i] = 50;> >>> >>         // smooth the cycle period
> >>> >>         Period[i] = 0.2 * Period[i] + 0.8 * Period[i-1];> >>> >>         SmoothPeriod[i] = 0.33 * Period[i] + 0.67 * SmoothPeriod[i-> > 1];> >>
> >>         }> >>> >>         Plot( SmoothPeriod, "Dominant Cycle", colorWhite,> > styleLine|styleOwnScale );> >>> >>         //Plot( Re, "Re", colorBlue, styleLine|styleOwnScale );
> >>> >>         //Plot( Im, "Im", colorSkyblue,styleLine|styleOwnScale );> >>> >>         //Plot( Im/Re, "Im/Re", colorDarkGreen,> > styleLine|styleOwnScale );
> >>> >>         //Plot( atan(Im/Re), "atan(Im/Re)", colorBrightGreen,> > styleLine|styleOwnScale );> >>> >>         //Plot( Period, "Period", colorYellow,
> > styleLine|styleOwnScale );> >>> >>> >>> >> -------------------------------------------------------------------> > -----> >>> >
> >> >> >> >> >> > Please note that this group is for discussion between users only.> >> > To get support from AmiBroker please send an e-mail directly to
> > SUPPORT {at} amibroker.com> >> > For other support material please check also:> > http://www.amibroker.com/support.html
> >> >> > Yahoo! Groups Links> >> >> >> >> >> >> >> >> >> >>
Please note that this group is for discussion between users only.To get support from AmiBroker please send an e-mail directly toSUPPORT {at} amibroker.comFor other support material please check also:
http://www.amibroker.com/support.htmlYahoo! Groups Links<*> To visit your group on the web, go to:    
http://groups.yahoo.com/group/amibroker/<*> Your email settings:    Individual Email | Traditional<*> To change settings online go to:    
http://groups.yahoo.com/group/amibroker/join    (Yahoo! ID required)<*> To change settings via email:    mailto:[EMAIL PROTECTED]
    mailto:[EMAIL PROTECTED]<*> To unsubscribe from this group, send an email to:    
[EMAIL PROTECTED]<*> Your use of Yahoo! Groups is subject to:    http://docs.yahoo.com/info/terms/


__._,_.___





Please note that this group is for discussion between users only.

To get support from AmiBroker please send an e-mail directly to 
SUPPORT {at} amibroker.com

For other support material please check also:
http://www.amibroker.com/support.html










   



    


  
  
    SPONSORED LINKS
  
          
                  
            
        Software support
      
                      
        Small business finance
      
                      
        Business finance online
      
              
                        
            
        Business finance training
      
                      
        Business finance course
      
                    
           
  






      
      Your email settings: Individual Email|Traditional 
      Change settings via the Web (Yahoo! ID required) 
      Change settings via email: Switch delivery to Daily Digest | Switch to Fully Featured 
           
        Visit Your Group 
       |
      
        Yahoo! Groups Terms of Use
       |
      
       Unsubscribe 
       
 
    
  




__,_._,___