META TRADER 4 THREAD NOW ON KRESLIK.COM

Moderator: moderators

User avatar
TheRumpledOne
rank: 10000+ posts
rank: 10000+ posts
Posts: 15544
Joined: Sun May 14, 2006 9:31 pm
Reputation: 3035
Location: Oregon
Real name: Avery T. Horton, Jr.
Gender: None specified
Contact:

Postby TheRumpledOne » Thu Dec 27, 2007 7:52 pm

Image

I turned the xMeter into a chart so you can see the history of the strength/weakness of the currencies.

Since this uses bid/ask info, when you load it up, it will NOT plot any history but will start from the load time. Just leave it running and the chart will fill out.
Attachments
TRO_xMeter.zip
(7.34 KiB) Downloaded 313 times
Last edited by TheRumpledOne on Mon Apr 07, 2008 1:06 pm, edited 1 time in total.
IT'S NOT WHAT YOU TRADE, IT'S HOW YOU TRADE IT!

Please do NOT PM me with trading or coding questions, post them in a thread.

Please add www.kreslik.com to your ad blocker white list.
Thank you for your support.

Poocher
rank: <50 posts
rank: <50 posts
Posts: 4
Joined: Tue Dec 18, 2007 8:10 pm
Reputation: 0
Location: Los Angeles
Gender: Male

Postby Poocher » Fri Dec 28, 2007 12:05 am

Well its runnin'.

I like squiggly lines but this sucker squiggles a lot!

User avatar
TheRumpledOne
rank: 10000+ posts
rank: 10000+ posts
Posts: 15544
Joined: Sun May 14, 2006 9:31 pm
Reputation: 3035
Location: Oregon
Real name: Avery T. Horton, Jr.
Gender: None specified
Contact:

Postby TheRumpledOne » Fri Dec 28, 2007 8:19 am

More MT4 code and updates.
Attachments
MT4_2007Dec27.zip
(246.41 KiB) Downloaded 363 times
IT'S NOT WHAT YOU TRADE, IT'S HOW YOU TRADE IT!



Please do NOT PM me with trading or coding questions, post them in a thread.

User avatar
TheRumpledOne
rank: 10000+ posts
rank: 10000+ posts
Posts: 15544
Joined: Sun May 14, 2006 9:31 pm
Reputation: 3035
Location: Oregon
Real name: Avery T. Horton, Jr.
Gender: None specified
Contact:

Postby TheRumpledOne » Fri Dec 28, 2007 7:18 pm

Image

MACD meter.
Attachments
tro_macd.zip
(24.67 KiB) Downloaded 303 times
IT'S NOT WHAT YOU TRADE, IT'S HOW YOU TRADE IT!



Please do NOT PM me with trading or coding questions, post them in a thread.

User avatar
TheRumpledOne
rank: 10000+ posts
rank: 10000+ posts
Posts: 15544
Joined: Sun May 14, 2006 9:31 pm
Reputation: 3035
Location: Oregon
Real name: Avery T. Horton, Jr.
Gender: None specified
Contact:

Postby TheRumpledOne » Fri Dec 28, 2007 9:59 pm

More MT4 code
Attachments
MT4_2007Dec28.zip
(209.39 KiB) Downloaded 340 times
IT'S NOT WHAT YOU TRADE, IT'S HOW YOU TRADE IT!



Please do NOT PM me with trading or coding questions, post them in a thread.

Please add www.kreslik.com to your ad blocker white list.
Thank you for your support.

User avatar
TheRumpledOne
rank: 10000+ posts
rank: 10000+ posts
Posts: 15544
Joined: Sun May 14, 2006 9:31 pm
Reputation: 3035
Location: Oregon
Real name: Avery T. Horton, Jr.
Gender: None specified
Contact:

Postby TheRumpledOne » Sat Dec 29, 2007 2:58 am

Image

I added the RSI METER and improved/fixed the MA and MACD meters.
Attachments
TRO_Meters.zip
(78.14 KiB) Downloaded 375 times
IT'S NOT WHAT YOU TRADE, IT'S HOW YOU TRADE IT!



Please do NOT PM me with trading or coding questions, post them in a thread.

richie
rank: <50 posts
rank: <50 posts
Posts: 12
Joined: Wed Oct 31, 2007 2:30 pm
Reputation: 0
Gender: None specified

Postby richie » Sat Dec 29, 2007 11:18 am

Hi, I am still a newbie in forex industry. All I can say, TRO you're sick man :D I wish have your brain as my EA.. thanks a lot for the indicators. Btw, do you trade for a living or just trade occasionally?

User avatar
TheRumpledOne
rank: 10000+ posts
rank: 10000+ posts
Posts: 15544
Joined: Sun May 14, 2006 9:31 pm
Reputation: 3035
Location: Oregon
Real name: Avery T. Horton, Jr.
Gender: None specified
Contact:

Postby TheRumpledOne » Sat Dec 29, 2007 3:05 pm

"TRO you're sick man"

You have no idea just how sick I am...LOL!
IT'S NOT WHAT YOU TRADE, IT'S HOW YOU TRADE IT!



Please do NOT PM me with trading or coding questions, post them in a thread.

fullyautomatix12
rank: <50 posts
rank: <50 posts
Posts: 2
Joined: Mon Dec 24, 2007 9:10 pm
Reputation: 0
Gender: Male

Postby fullyautomatix12 » Sat Dec 29, 2007 5:40 pm

TRO,

Please can you convert this ELD to MetaTrader?

I am no longer with TradeStation so cannot show you a snapshot of the indicator.

Code: Select all

{
   XCAP_iPolyCycle
    
   Author: Paul A. Griffin
   January 16, 2007
   Extrema Capital, 2007   
   
   Introduction:

   This is an example of what can be done by combining Legendre polynomials and
   analytic signals.  I get a way of determining a smooth period and
   relative adaptive strength indicator without adding time lag.

   This indicator displays the following:

   a. The Least Squares fit of a polynomial to a DC subtracted time series - a best fit to a cycle.

   b. The normalized analytic signal of the cycle (signal and quadrature).

   c. The Phase shift of the analytic signal per bar.

   d. The Period and HalfPeriod lengths, in bars of the current cycle.

   e. A relative strength indicator of the time series over the cycle length.  That is, adaptive
   relative strength over the cycle length.

   The Relative Strength Indicator, is adaptive to the time series, and it can be smoothed by increasing the length of
   decreasing the number of degrees of freedom.
   
   Other adaptive indicators based upon the period and can be similarly constructed. 
   
   There is some new math here, so I have broken the story up into 5 Parts:   
   
   Part 1:

   Any time series can be decomposed into a orthogonal set of polynomials [1,2,3]. 
   This is just math and here are some good references:
   
   [1] http://en.wikipedia.org/wiki/Legendre_polynomials

   [2] https://www.tradestation.com/Discussions/Topic.aspx?Topic_ID=59250 

   [3] Peter Seffen, "On Digital Smoothing Filters: A Brief Review of Closed Form Solutions and Two New Filter Approaches",
      Circuits Systems Signal Process, Vol. 5, No 2, 1986

   I gave some thought to what should be done with ths and came to the conclusion that they can be used for basic smoothing
   of time series.  For the analysis below, I decompose a time series into a low number of
   degrees of freedom and discard the zero mode to introduce smoothing.

   That is:

   time series => c_1 t + c_2 t^2 ... c_Max t^Max

   This is the cycle.  By construction, the cycle does not have a zero mode and more physically, I am defining the
    "Trend" to be the zero mode. 

   The data for the cycle and the fit of the cycle can be viewed by setting

   ShowDataAndFit = TRUE; 

   There, you will see the fit of the last bar as well as the time series of the leading edge of the fits.  If you don't
   know what I mean by the "leading edge", please see some of the postings in [2].  The leading edges are in grayscale,
   and the fit of the last bar is in color.

   I have choosen Length = 17 and Degree = 4 as the default.  I am simply making sure by eye that the fit is reasonably good and
   degree 4 is the lowest polynomial that can represent a sine-like wave, and 17 is the smallest length that lets me calculate
   the Phase Shift (Part 3 below) using the Hilbert Transform of width=7 (Part 2 below). 

   Depending upon the fit you make, you will capture different cycles in the data.  A fit that is too "smooth" will
    not see the smaller cycles, and a fit that is too "choppy" will not see the longer ones.  The idea is to use the fit
    to try to suppress the smaller noise cycles while keeping larger signal cycles. 

   Part 2:
 
   Every time series has an Analytic Signal, defined by applying the Hilbert Transform to it. You can think of the
   original time series as amplitude * cosine(theta) and the transformed series, called the quadrature, can be thought of as
   amplitude * sine(theta).  By taking the ratio, you can get the angle theta, and this is exactly what was done by
   John Ehlers in [4].  It lets you get a frequency out of the time series under consideration. 
   
   [4] http://www.amazon.com/Rocket-Science-Traders-Processing-Applications/dp/0471405671
   
   It helps to have more references to understand this.  There is a nice article in Wikipedia[5] on it. 
   Read the part about the discrete Hilbert Transform:
   
   [5] http://www.answers.com/topic/hilbert-transform

   Also, Answers.com has good information on the Hilbert Transform and links to other concepts:

   [6] http://www.answers.com/topic/hilbert-transform
   
   If you really want to understand how to go from continuous to discrete, look up this article
   written by Richard Lyons:

   [7] www.dspguru.com/info/tutor/QuadSignals.pdf

   In the indicator below, I am calculating the normalized analytic signal,
   which can be written as:

   s + i h where i is the imagary number, and s^2 + h^2 = 1;

   s= signal = cosine(theta)
   h = hilbert transformed signal = quadrature = sine(theta)

   The angle is therefore given by theta = arctan(h/s);

   The analytic signal leading edge and the fit of the last bar of the cycle can be viewed by setting

   ShowAnalyticSignal = TRUE; 
   
   The leading edges are in grayscale fit to the last bar is in color.  Light (yellow) is the
   s term, and Dark (orange) is the quadrature (hilbert transform).  Note that for every bar,
   s^2 + h^2 = 1 , by construction.

   I am using a width = 7 Hilbert transform, just like Ehlers.  (But you can adjust it if you want.)  This transform has a
   7 bar lag.  I have put the lag into the plot statements, so the cycle info should be quite good at displaying minima
   and maxima (extrema).
      
   Part 3:

   The Phase shift is the amount of phase change from bar to bar. 

   It is a discrete unitary transformation that takes s[1] + i h[1] to  s + i h

   explicitly, T = (s+ih)*(s[1]-ih[1]) , since s[1]*s[1] + h[1]*h[1] = 1.

    writing it out, we find that T =  T1 + iT2

    where T1 = s*s[1] + h*h[1] and T2 = s*h[1]-h*s[1]
   
   and the phase shift is given by PhaseShift = arctan(T2/T1); 

   Alas, I have no reference for this, all I doing is finding the rotation what takes the analytic signal
   at bar [1] to the analytic signal at bar [0].  T is the transfer matrix.

   Of interest is the PhaseShift from the closest two bars to the present, given by the
   bar [7] and bar [8] since I am using a width=7 Hilbert transform, bar [7] is the earliest bar with an
   analytic signal.

   I store the phase shift from bar [7] to bar [8] as a timeseries called PhaseShift.  It basically gives
   you the (7-bar delayed) leading edge the amount of phase angle change in the series.

   You can see it by setting

   ShowPhaseShift=TRUE

   The green points are positve phase shifts and red points are negative pahse shifts.
    On most charts, I have looked at, the indicator is mostly green, but occationally, the stock
   "retrogrades" and red appears.  This happens when the cycle is "broken" and the cycle length starts to
   expand as a trend occurs.
 
   Part 4:

   The Period:

   The Period is the number of bars required to generate a sum of
   PhaseShifts equal to 360 degrees.   
   
   The halfperiod is the number of bars required to generate a sum of phase shifts
   equal to 180 degrees. It is usually not equal to 1/2 of the period.

   You can see the Period and Halfperiod by setting

   ShowPeriod=TRUE

   The code is very simple here:

   Value1=0;
   Value2=0;
   while Value1 < barnumber and AbsValue(Value2) < 360 begin
      Value2 = Value2 + PhaseShift[Value1];
      Value1 = Value1 + 1;
   end;
   Period = Value1;

   The period is sensitive to the input length and degree values but not overly so.  Any insight
   on this would be appreciated.

   Part 5:

   The Relative Stength indicator:

   The Relative Strength is just the current value of the series minus the minimum over the last cycle
   divided by the maximum - minimum over the last cycle, normalized between +1 and -1.

   RelativeStrength = -1 + 2*(Series-Min)/(Max-Min);

   It therefore tells you where the current bar is relative to the cycle.  If you want to smooth the indicator, then
   extend the period and/or reduce the polynomial degree.

   In code:
   
   NewLength = floor(Period + HilbertWidth+1);
   Max = highest(Series,NewLength);
   Min = lowest(Series,NewLength);
   if Max>Min then

   Note that the variable NewLength includes the lag that comes from the hilbert transform,
   (HilbertWidth=7 by default).

   Conclusion:

   This is an example of what can be done by combining Legendre polynomials and
   analytic signals to determine a smooth period without adding time lag.

}

Inputs:   
   Series((h+l)/2),
   Length(17),               // Length = 2*Width+1, the minimum is 17 bars
   Degree(4),               // 0 <= Degree <= 2*Width+1, default is 4, 4 DOF for 17 bars
   ShowDataAndFit(FALSE),      //Shows data and fit of the data last bar only
   ShowAnalyticSignal(FALSE),   //Shows the signal between the
    ShowPhaseShift(FALSE),      //Shows the PhaseShift for the last bar
   ShowPeriod(FALSE),         //Shows the number of bars required to Phase Shift a half and a full period
   ShowRelativeStrength(TRUE);   //Show the relative strengh of the current bar w.r.t. a full period (+ calculation lag). 

 Variables:
    HilbertWidth(7),         // Must be an odd number.  1,3,5,7,9, ... less than or equal to Width = floor(Length-1)/2)
   pi(3.141592653589793),      //close enough
   Width(floor((Length-1)/2));   //      

  Variables:
     DC(0),
     EarliestWidth(0),
      LeadingCycleEdge(0),
      LeadingSignalEdge(0),
   LeadingHilbertEdge(0),
      p(0),j(0),k(0),
      DataSize(2*Width+1),
   T1(0),T2(0),//TransferMatrix
   PhaseShift(0),
   HalfPeriod(0),Period(0),
     Max(0),Min(0),
   RelativeStrength(0);


Arrays:
   Polynomial[](0),
   Coefficient[](0),
   Cycle[](0),
   Signal[](0),
   Hilbert[](0),
   HilbertTransform[](0),
   SignalTransform[](0);

 if barnumber = 1 then begin

   //Allocate memory for the arrays.  Polynomial is a 2D array
   Array_SetMaxIndex(Polynomial, DataSize*(Degree+1)+1);
   Array_SetMaxIndex(Coefficient, Degree+1);
   Array_SetMaxIndex(Cycle, DataSize);
   Array_SetMaxIndex(Signal, DataSize);
   Array_SetMaxIndex(Hilbert, DataSize);

   //We are going to make the analytic signal of the time series with the transforms below:
   Array_SetMaxIndex(HilbertTransform, 2*HilbertWidth+1);
   Array_SetMaxIndex(SignalTransform,2*HilbertWidth+1);

   //Create the Dicrete Hilbert Transform Filter, normalized to a step function (the last being my idea)
   for k = -HilbertWidth to HilbertWidth begin   
      if mod(k,2) = 0 then HilbertTransform[HilbertWidth+k] = 0 else HilbertTransform[HilbertWidth+k] = 2/(pi*k);      
      if k = 0 then SignalTransform[HilbertWidth+k] = 1 else SignalTransform[HilbertWidth+k] = 0;
   end;
      
end;

//Determine earliest bar required for analysis and plotting to save time
if LastBarOnChart = TRUE then EarliestWidth = - Width else EarliestWidth = Width - 2*HilbertWidth-1;

/////////////////////////////////////////////////////////////////////
//   Decompose data into the polynomial set and determine the overlap coefficients:
//   
//   This is the basic decomposition step
   for p = 0 to Degree begin
      Coefficient[p] = 0;
       for j = -Width to  + Width begin
         Coefficient[p]=   Coefficient[p] + Polynomial[p*DataSize+Width+j]*Series[Width-j];
      end;   
   end;
//
/////////////////////////////////////////////////////////////////////

// Remap to the Cycle to create the Cycle time series, remove the 0 mode, and the modes higher than Degree.
//
   DC = Coefficient[0]*Polynomial[Width+Width];
   
   for j  = EarliestWidth to Width begin         
      Cycle[Width+j] = 0;   
      for p = 1 to Degree begin //p=0 removed
         Cycle[Width+j] = Cycle[Width+j] + Coefficient[p]*Polynomial[p*DataSize+Width+j];
      end;
   end;
//
////////////////////////////////////////////////////////////////////////


/////////////////////////////////////////////////////////////////////////
// Signal and Hilbert Of the Cycle: Create the Analytic time series
//
   for j  = EarliestWidth + HilbertWidth to Width - HilbertWidth begin               
      
      Signal[Width+j]=0;
      Hilbert[Width+j] = 0;
   
      //Execute the Transforms
      for k = -HilbertWidth to HilbertWidth begin   
         Hilbert[Width+j] = Hilbert[Width+j] + HilbertTransform[HilbertWidth+k] * Cycle[Width+j+k];
         Signal[Width+j] = Signal[Width+j] + SignalTransform[HilbertWidth+k] * Cycle[Width+j+k];
      end;
   
      //The normalized with respect to the amplitude
      Value1 = Power(Signal[Width+j],2) +  Power(Hilbert[Width+j],2);
      if Value1>0 then begin
         Value1 = Power(Value1,-0.5);
         Signal[Width+j]= Value1 * Signal[Width+j];
         Hilbert[Width+j]= Value1 * Hilbert[Width+j];
      end;
   
   end;
//
//////////////////////////////////////////////////////////////////////////


//////////////////////////////////////////////////////////////////////////
// The Phase Shift of the leading edge bar change, by Paul Griffin
   
   T1 = Signal[2*Width-HilbertWidth]*Signal[2*Width-HilbertWidth-1] +
      + Hilbert[2*Width-HilbertWidth]*Hilbert[2*Width-HilbertWidth-1];
   
   T2 = -Hilbert[2*Width-HilbertWidth]*Signal[2*Width-HilbertWidth-1] 
      + Signal[2*Width-HilbertWidth]*Hilbert[2*Width-HilbertWidth-1];
      
   If T1 <> 0  then PhaseShift = Arctangent(T2 / T1);
//
//////////////////////////////////////////////////////////////////////
   
//////////////////////////////////////////////////////////////////////
//  Add phase shifts to determine the period and halfperiod, by Paul Griffin
   
   //period
   Value1=0;
   Value2=0;
   while Value1 < barnumber and AbsValue(Value2) < 360 begin
      Value2 = Value2 + PhaseShift[Value1];
      Value1 = Value1 + 1;
   end;
   Period = Value1;

   //halfperiod
   Value1=0;
   Value2=0;
   while Value1 < barnumber and AbsValue(Value2) < 180 begin
      Value2 = Value2 + PhaseShift[Value1];
      Value1 = Value1 + 1;
   end;
   HalfPeriod = Value1;

   //Store the leading edges for display purposes
   LeadingSignalEdge =  Signal[Width+Width - HilbertWidth];
   LeadingHilbertEdge =  Hilbert[Width+Width - HilbertWidth];
   LeadingCycleEdge =  Cycle[Width+Width];

//////////////////////////////////////////////////////////////////
// The Relative Strength, this is the easy part

   Value1 = floor(Period + HilbertWidth+1);
   Max = highest(Series,Value1);
   Min = lowest(Series,Value1);
   if Max>Min then RelativeStrength = -1 + 2*(Series-Min)/(Max-Min);
//
//////////////////////////////////////////////////////////////////
//////////////////////////////////////////////////////////////////
//
//   Plotting routines
      
   if LastBarOnChart = TRUE and ShowDataAndFit = TRUE then begin
      for j  = EarliestWidth to Width begin
         plot1[Width-j](Series[Width-j]-DC,"Data",cyan);
         plot2[Width-j](Cycle[Width+j],"Cycle",blue);
      end;
   end;

   if barnumber>1 and ShowDataAndFit = TRUE then begin
      plot3(Series-DC,"LeadingData",lightgray);
      plot4(LeadingCycleEdge,"LeadingCycle",white);
   end;

   if LastBarOnChart = TRUE and ShowAnalyticSignal = TRUE then begin
      for j  = -Width + HilbertWidth to Width - HilbertWidth begin
         plot5[Width-j](Signal[Width+j],"Signal",yellow);
         plot6[Width-j](Hilbert[Width+j],"Quadrature",RGB(255,128,0));
      end;
   end;

   if barnumber>1 and  ShowAnalyticSignal = TRUE then begin
      Plot7[HilbertWidth](LeadingSignalEdge,"SignalEdge",lightgray);
      Plot8[HilbertWidth](LeadingHilbertEdge,"QuadEdge",darkgray);
   end;

   if barnumber>1 and ShowPhaseShift = TRUE then begin
      if T2>0 then Plot9[HilbertWidth](PhaseShift,"PhaseShift",green);
      if T2<0 then Plot9[HilbertWidth](PhaseShift,"PhaseShift",red);
   end;
   if barnumber>1 and ShowPeriod=TRUE then begin
      Plot10[HilbertWidth](Period,"Period",yellow);
      Plot11[HilbertWidth](HalfPeriod,"HalfPeriod",RGB(255,128,0));
   end;

   if barnumber>1 and ShowRelativeStrength=TRUE then begin
      plot12(RelativeStrength,"RelStrength",white);
   end;

   Plot99(0,"0",white);
//
///////////////////////////////////////////////////////////////
      
///////////////////////////////////////////////////////////////
//
//   Create the Orthoginal Legendre Polynomials during the first bar
//
////////////////////////////////////////////////////////////////   
if barnumber = 1 then begin

//The first polynomial is a constant, easy to create
for j = - Width to + Width begin
   Polynomial[0*DataSize + Width+j] = 1;
end;

//The second polynomial is a line
if Degree>=1 then begin
   for j = -Width to +Width  begin
      Polynomial[1*DataSize + Width+j] = j;
   end;
end;

//We use the discrete Legendre polynomial recurrence relations to create the rest
if Degree > 1 then begin      
   for p = 1 to Degree - 1 begin                        // create the polynomial for degree p+1      
      for j = -Width to Width  begin                        // sum over the interval
         
         Value1 =    j*(2*p+1)/(p+1);                     //discrete Legendre polynomial solution
         Value2 =     - (p/(p+1))*(2*Width+1+p)*(2*Width+1-p)/4;    //discrete Legendre polynomial solution

         Polynomial[(p+1)*DataSize+Width+j]                   //The recurrence relation
         = Value1 * Polynomial[p*DataSize+Width+j] + Value2 * Polynomial[(p-1)*DataSize+Width+j];
      end;
   end;
end;

//Now we have to normalize each polynomial, so that the sum of the square of the polynomial
//over the interval is 1. Instead of doing the calculation however, we apply the pre-determined answer[4]:
for p = 0 to Degree begin      
   Value1 = Power(2,-2*p)/(2*p+1);
   for j = -p to p begin Value1 = Value1 * (2*Width+1+j); end;
   if Value1 > 0 then Value1 = 1/squareroot(Value1);  //this is the normalization coefficient

    for j = -Width to +Width  begin
      Polynomial[p*DataSize+Width+j] = Value1 * Polynomial[p*DataSize+Width+j];
   end;
end;

//Done!  We now have a orthogonal normalized set of polynomials
      
end;


Thanks much for your help.

User avatar
TheRumpledOne
rank: 10000+ posts
rank: 10000+ posts
Posts: 15544
Joined: Sun May 14, 2006 9:31 pm
Reputation: 3035
Location: Oregon
Real name: Avery T. Horton, Jr.
Gender: None specified
Contact:

Postby TheRumpledOne » Sat Dec 29, 2007 6:05 pm

NO.

I am not programming other people's code into MT4.

I am only programming the tools necessary for my style of trading.

I am sure you can find something like the code you want. There are THOUSANDS of free MT4 programs on the net.
IT'S NOT WHAT YOU TRADE, IT'S HOW YOU TRADE IT!



Please do NOT PM me with trading or coding questions, post them in a thread.

Please add www.kreslik.com to your ad blocker white list.
Thank you for your support.


Return to “MetaTrader”