refactoring

LAB2_2/Assignment1/hopfield.m

@@ -1,59 +1,43 @@
 % Hopfield model
-% <https://elearning.di.unipi.it/pluginfile.php/46092/mod_resource/content/0/LAB2_2-Assignments.pd>
+% <https://elearning.di.unipi.it/mod/resource/view.php?id=10695>
 
 load('lab2_2_data');
 set(groot,'defaultFigureVisible','off');
 
 % distorted images of pattern 0
-d_0_1_orig = distort_image(p0,0.05);
-d_0_2_orig = distort_image(p0,0.1);
-d_0_3_orig = distort_image(p0,0.25);
-d_0_1 = d_0_1_orig;
-d_0_2 = d_0_2_orig;
-d_0_3 = d_0_3_orig;
+d_0_1 = distort_image(p0,0.05); d_0_2 = distort_image(p0,0.1); d_0_3 = distort_image(p0,0.25);
 % distorted images of pattern 1
-d_1_1_orig = distort_image(p1,0.05);
-d_1_2_orig = distort_image(p1,0.1);
-d_1_3_orig = distort_image(p1,0.25);
-d_1_1 = d_1_1_orig;
-d_1_2 = d_1_2_orig;
-d_1_3 = d_1_3_orig;
+d_1_1 = distort_image(p1,0.05); d_1_2 = distort_image(p1,0.1); d_1_3 = distort_image(p1,0.25);
 % distorted images of pattern 2
-d_2_1_orig = distort_image(p2,0.05);
-d_2_2_orig = distort_image(p2,0.1);
-d_2_3_orig = distort_image(p2,0.25);
-d_2_1 = d_2_1_orig;
-d_2_2 = d_2_2_orig;
-d_2_3 = d_2_3_orig;
-%imagesc([reshape(d_0_1,32,32) reshape(d_0_2,32,32) reshape(d_0_3,32,32)]);
+d_2_1 = distort_image(p2,0.05); d_2_2 = distort_image(p2,0.1); d_2_3 = distort_image(p2,0.25);
 eps = 1;
 
-
-% (1) storage phase (learning)
+%% (1) storage phase (learning)
 P = [p0';p1';p2'];
 W = 1/1024 * (P' * P);
 for i=1:1024
     W(i,i) = 0;
 end
 I = ones(1024,1)*0.5;  % bias
 
-% (2) retrieval phase
-% (2.1) initialization
-O = [d_0_1_orig';d_0_2_orig';d_0_3_orig';d_1_1_orig';d_1_2_orig';d_1_3_orig';d_2_1_orig';d_2_2_orig';d_2_3_orig'];  % original images
-D = [d_0_1';d_0_2';d_0_3';d_1_1';d_1_2';d_1_3';d_2_1';d_2_2';d_2_3'];  % distorted images
+%% (2) retrieval phase
+%% (2.1) initialization
+O = [d_0_1';d_0_2';d_0_3';d_1_1';d_1_2';d_1_3';d_2_1';d_2_2';d_2_3'];  % original images
+D = O;  % distorted images
 L = [[0,0.05];[0,0.10];[0,0.25];[1,0.05];[1,0.10];[1,0.25];[2,0.05];[2,0.10];[2,0.25]];  % patterns info
-Ps = [p0';p0';p0';p1';p1';p1';p2';p2';p2'];
+Ps = [p0';p0';p0';p1';p1';p1';p2';p2';p2'];  % patterns of each distorted images
 
+% for each pattern distortions
 for p=1:size(D,1)
     epoch = 1;
     overlaps = [];
     energies = [];
     pattern_num = L(p,1);
     pattern_distortion = L(p,2);
     energy_old = energy(W, I, D(p,:));
-    % (2.2) iteration until convergence
+    %% (2.2) iteration until convergence
     while 1
-        % (2.3) choose i random neuron
+        %% (2.3) choose i random neuron
         for i=randperm(1024)
             D(p,i) = sign(W(i,:) * D(p,:)' + I(i));
             overlaps = horzcat(overlaps,[overlap(p0, D(p,:)) overlap(p1, D(p,:)) overlap(p2, D(p,:))]');
@@ -69,22 +53,23 @@
         epoch = epoch + 1;
     end
 
+    %% plottings
     disp('Saving figures . . .');
     fig = figure;
     hold on;
     plot((1:size(overlaps,2)),overlaps(1,:));
     plot((1:size(overlaps,2)),overlaps(2,:));
     plot((1:size(overlaps,2)),overlaps(3,:));
     title(sprintf('Overlaps (pattern=%d, distortion=%0.2f)',pattern_num,pattern_distortion));
-    xlabel('ticks')
+    xlabel('ticks');
     ylabel('overlap function');
-    legend('overlap with pattern 0','overlap with pattern 1','overlap with pattern 2');
+    legend({'overlap with pattern 0','overlap with pattern 1','overlap with pattern 2'},'Location','southeast');
     print(fig,sprintf('images/distorted_%d_%0.2f_overlap.png',pattern_num,pattern_distortion),'-dpng');
     hold off;
     fig = figure;
     plot((1:size(energies,2)),energies(1,:));
     title(sprintf('Energy (pattern=%d, distortion=%0.2f)',pattern_num,pattern_distortion));
-    xlabel('ticks')
+    xlabel('ticks');
     ylabel('energy');
     print(fig,sprintf('images/distorted_%d_%0.2f_energy.png',pattern_num,pattern_distortion),'-dpng');
     fig = figure;
@@ -93,7 +78,6 @@
     print(fig,sprintf('images/distorted_%d_%0.2f_reconstructed.png',pattern_num,pattern_distortion),'-dpng');
 end
 
-
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
 function m = overlap(p, x)
     m = (p' * x') / 1024;