Entropy Indicator in TradeStation using Matlab: TradeStation TS.Entropy Indicator
Here is an example of computation of entropy in TradeStation using matlab
| EasyLanguage: | {*******************************************************************
Name: TS.MATLAB.Entropy
Analysis Type: Indicator
Description: An example of using Matlab into Tradeststion
Used: TS.MATLAB.LINK.dll
Provided By: Trade Smart Research (c) Copyright 2001 - 2005
www.tsresearch.com
*******************************************************************}
Input: Length(500), q(1);
{create array x and add new value}
if currentbar = 1 then begin
Value1 = TS.MATLAB.LINK("x = ["+ NumToStr(C,4)+" ]");
end else begin
{add next value to array}
Value1=TS.MATLAB.LINK("x=[x "+ NumToStr(C,4) +" ]");
if currentbar > Length then
{if array > our periods (length) then delete first element from array}
Value1=TS.MATLAB.LINK("x(1)=[]");
{calculating TSentropy.m function}
Value1=TS.MATLAB.LINK("TSentropy(x,"+NumToStr(q,2)+")");
end;
Plot1(value1, "Entropy");
{***** Copyright (c) 2001-2005 Trade Smart Research, Ltd. All rights reserved. www.tsresearchgroup.com *****
***** Trade Smart Research reserves the right to modify or overwrite this analysis technique
with each release. *****} |
| Matlab: | function entr = TSentropy(x, q)
% TSentropy estimates the entropy of signals .
% entr : The entropy estimate
% q : input parameter, q >=1;
% x : The time series to be analyzed
% q : Tsallis non-extensive parameter value, q >=1;
% if q=1 then Tsallis' entropy concides with Shannon's
% http://www.TSResearch.com
% D. Tolstonogov
% Copyright (c) by Trade Smart Research
% 08/04/2004
[NRow,NCol]=size(x);
if NRow~=1
x=x';
end;
% The number of cells of the histogram
minx=min(x);
maxx=max(x);
delta=(maxx-minx)/(length(x)-1);
ncell=ceil(sqrt(length(x)));
% histogram
h = hist(x, ncell);
entr=0;
count=0;
% Shannon's entropy
if q==1
for n=1:ncell
if h(n)~=0
hn=log(h(n));
else
hn=0;
end;
count=count+h(n);
entr=entr-h(n)*hn;
end;
entr=entr/count+log(count);
else
% Tsallis entropy
for n=1:ncell
hn=(h(n))^q;
count=count+h(n);
entr=entr+hn;
end;
entr=entr/(count)^q;
entr=(1-entr)/(q-1);
end; |
|
