fix demo in chapter03

chapter01/demo.m

@@ -1,4 +1,6 @@
-
+% Done
+% demo for information theory toolbox
+clear;
 k = 10;  % variable range
 n = 100;  % number of variables
 
@@ -12,27 +14,27 @@
 Hx = entropy(x);
 Hy = entropy(y);
 %% joint entropy H(x,y)
-Hxy = jointEntropy(x, y);
+Hxy = jointEntropy(x,y);
 %% conditional entropy H(x|y)
-Hx_y = condEntropy(x, y);
+Hx_y = condEntropy(x,y);
 %% mutual information I(x,y)
-Ixy = mutInfo(x, y);
+Ixy = mutInfo(x,y);
 %% relative entropy (KL divergence) KL(p(x)|p(y))
-Dxy = relatEntropy(x, y);
+Dxy = relatEntropy(x,y);
 %% normalized mutual information I_n(x,y)
-nIxy = nmi(x, y);
+nIxy = nmi(x,y);
 %% nomalized variation information I_v(x,y)
-vIxy = nvi(x, y);
+vIxy = nvi(x,y);
 %% H(x|y) = H(x,y)-H(y)
-isequalf(Hx_y, Hxy-Hy)
+isequalf(Hx_y,Hxy-Hy)
 %% I(x,y) = H(x)-H(x|y)
-isequalf(Ixy, Hx-Hx_y)
+isequalf(Ixy,Hx-Hx_y)
 %% I(x,y) = H(x)+H(y)-H(x,y)
-isequalf(Ixy, Hx+Hy-Hxy)
+isequalf(Ixy,Hx+Hy-Hxy)
 %% I_n(x,y) = I(x,y)/sqrt(H(x)*H(y))
-isequalf(nIxy, Ixy/sqrt(Hx*Hy))
+isequalf(nIxy,Ixy/sqrt(Hx*Hy))
 %% I_v(x,y) = (1-I(x,y)/H(x,y))
-isequalf(vIxy, 1-Ixy/Hxy)
+isequalf(vIxy,1-Ixy/Hxy)
 
 
 

chapter03/demo.m

@@ -1,15 +1,20 @@
 % Done
 % demo for chapter 03
 clear; close all;
-n = 100;
-beta = 1e-1;
-X = rand(1,n);
-w = randn;
-b = randn;
-t = w'*X+b+beta*randn(1,n);
-x = linspace(min(X)-1,max(X)+1,n);   % test data
+% n = 100;
+% beta = 1e-1;
+% X = rand(1,n);
+% w = randn;
+% b = randn;
+% t = w'*X+b+beta*randn(1,n);
+% x = linspace(min(X)-1,max(X)+1,n);   % test data
+
+d = 1;
+n = 200;
+[X,t,model] = linRnd(d,n);
+plotBand(X,t,2*model.sigma);
 %%
-model = regress(X, t);
+model = linReg(X,t);
 y = linInfer(x, model);
 figure;
 hold on;
@@ -18,7 +23,7 @@
 hold off
 pause
 %%
-[model,llh] = regressEbEm(X,t);
+[model,llh] = linRegEbEm(X,t);
 [y, sigma] = linInfer(x,model,t);
 figure;
 hold on;
@@ -30,7 +35,7 @@
 plot(llh);
 pause
 %%
-[model,llh] = regressEbFp(X,t);
+[model,llh] = linRegEbFp(X,t);
 [y, sigma] = linInfer(x,model,t);
 figure;
 hold on;

chapter03/linRnd.asv

@@ -0,0 +1,19 @@
+function [X, t, model] = linRnd(d, n, prior)
+% Generate a data from a linear model p(t|w,x)=G(w'x+w0,sigma), where w and w0 are
+% generated from G(0,
+%   d: dimension of data
+%   n: number of data
+%   prior: a structure specify the prior
+
+if nargin < 3
+    sigma = 1;
+else
+    sigma = prior.sigma;
+end
+X = rand(d,n);
+w = randn(d,1);
+w0 = randn(1,1);
+epsilon = sigma^2*randn(1,n);
+t = w'*X+w0+
+model.w = w;
+model.w0 = w0;

chapter03/linRnd.m

@@ -0,0 +1,20 @@
+function [X, t, model] = linRnd(d, n, prior)
+% Generate a data from a linear model p(t|w,x)=G(w'x+w0,sigma), where w and w0 are
+% generated from G(0,1)
+%   d: dimension of data
+%   n: number of data
+%   prior: a structure specify the prior
+
+if nargin < 3
+    sigma = 1;
+else
+    sigma = prior.sigma;
+end
+X = rand(d,n);
+w = randn(d,1);
+w0 = randn(1,1);
+epsilon = sigma^2*randn(1,n);
+t = w'*X+w0+epsilon;
+model.w = w;
+model.w0 = w0;
+model.sigma = sigma;