Related to mutual information

This is a function for mutual information

function I = MutualInformation(X,Y);

if (size(X,2) > 1)  % More than one predictor?
    % Axiom of information theory
    I = JointEntropy(X) + Entropy(Y) - JointEntropy([X Y]);
else
    % Axiom of information theory
    I = Entropy(X) + Entropy(Y) - JointEntropy([X Y]);
end

But while running it is showing "not enough input arguements". I am not getting the problem. Can anyone please help me?

[ye song]

you should put X&Y

Answer 1

[Walter Roberson]

You need to go to the command line and call the routine, such as

   MutualInformation(rand(5,7), rand(5, 14))

[Anamika]

I am not getting what you want to mean? Can you please explain it?

[Walter Roberson]

You are trying to run your program by pressing F5 or by using the menu Run button. When you do that no argument will be passed into your routine. You need to instead go to the command line and call upon the routine like you would if you were using it inside a program, passing in your X and Y arrays. I used rand() to generate X and Y data for the purposes of illustration, but you should use data appropriate for you.

[Anamika]

Thanks a lot sir. I want to explain you one thing. First I have created the function for entropy. The function is shown below-

function h = entropy(vec1)
%=========================================================
%
%This is a prog in the MutualInfo 0.9 package written by 
% Hanchuan Peng.
%
%Disclaimer: The author of program is Hanchuan Peng
%      at <[e-mail]> and <[e-mail]>.
%
%The CopyRight is reserved by the author.
%
%Last modification: April/19/2002
%
%========================================================
%
% h = entropy(vec1)
% calculate the entropy of a variable vec1
%
% demo: 
%  a=[1 2 1 2 1]';b=[2 1 2 1 1]';
%  fprintf('entropy(a) = %d\n',entropy(a));
%
% the same as entropycond(vec1)
%
% By Hanchuan Peng, April/2002
%

if nargin<1,

    disp('Usage: h = entropy(vec1).');
    h = -1;

else,

    [p1] = estpa(vec1);
    h = estentropy(p1);

end;

Next I am generating the joint entropy function as follows-

function h = jointentropy(vec1,vec2)
%=========================================================
%
%This is a prog in the MutualInfo 0.9 package written by 
% Hanchuan Peng.
%
%Disclaimer: The author of program is Hanchuan Peng
%      at <[e-mail]> and <[e-mail]>.
%
%The CopyRight is reserved by the author.
%
%Last modification: April/19/2002
%
%========================================================
%
% h = jointentropy(vec1,vec2)
% calculate the joint entropy of two variables
%
% when only one variable presents, this function equals entropy(vec1)
%
% demo: 
%  a=[1 2 1 2 1]';b=[2 1 2 1 1]';
%  fprintf('jointentropy(a,b)= %d \n',jointentropy(a,b));
%
% By Hanchuan Peng, April/2002
%

if nargin<1,

    disp('Usage: h = condentropy(vec1,<vec2>).');
    h = -1;

elseif nargin<2,

    [p1] = estpa(vec1);
    h = estentropy(p1);

else,

    [p12] = estpab(vec1,vec2);
    h = estjointentropy(p12);

end;

Next I am doing the above written mutual information function. Now I am writing one different program to call the function as-

clc;
clear all;
close all;
p=MutualInformation(rand(5,4),rand(5,14));

But when I run the program it is showing that "undefined function estpafor character double". Why this problem is creating?

[Walter Roberson]

The code calls upon the routine named "estpa", but you do not have code for that routine.