Plot each MATLAB quality history using its actual iteration count

MATLAB/Demos/d10_QualityMeasurements.m

@@ -103,29 +103,24 @@
 
 %% Plot the changes of image quality measures over iterations
 
-%%RMSE plot
-figure
-plot([qualitySIRT(1,:);[qualitySART(1,:) nan(1,length(qualityOSSART)-length(qualitySART))];qualityOSSART(1,:)]');
-title('Evolution of RMSE per iteration')
-legend('SIRT','SART','OS-SART')
-
-%%CC plot
-figure
-plot([qualitySIRT(2,:);[qualitySART(2,:) nan(1,length(qualityOSSART)-length(qualitySART))];qualityOSSART(2,:)]');
-title('Evolution of correlation coefficient (CC) per iteration')
-legend('SIRT','SART','OS-SART')
-
-%%MSSIM plot
-figure
-plot([qualitySIRT(3,:);[qualitySART(3,:) nan(1,length(qualityOSSART)-length(qualitySART))];qualityOSSART(3,:)]');
-title('Evolution of mean structural similarity index per iteration')
-legend('SIRT','SART','OS-SART')
-
-%%UQI plot
-figure
-plot([qualitySIRT(4,:);[qualitySART(4,:) nan(1,length(qualityOSSART)-length(qualitySART))];qualityOSSART(4,:)]');
-title('Evolution of universal quality index per iteration')
-legend('SIRT','SART','OS-SART')
+%%RMSE plot (and the other requested metrics)
+% Each algorithm may stop at a different iteration. Plot its own history
+% directly, preserving legitimate zeros and avoiding assumptions about length.
+histories={qualitySIRT,qualitySART,qualityOSSART};
+algorithmNames={'SIRT','SART','OS-SART'};
+metricNames={'RMSE','CC','MSSIM','UQI'};
+for metric=1:numel(metricNames)
+    figure;
+    hold on;
+    for algorithm=1:numel(histories)
+        values=histories{algorithm}(metric,:);
+        plot(1:numel(values),values,'DisplayName',algorithmNames{algorithm});
+    end
+    title(['Evolution of ' metricNames{metric} ' per iteration']);
+    xlabel('Iteration');
+    ylabel(metricNames{metric});
+    legend('show');
+end
 
 %% Compute the parameters for the target image
 % Knowing change of the parameters per iteration can be interesting, but

MATLAB/Tests/test_qualityDemoHistories.m

@@ -0,0 +1,43 @@
+function test_qualityDemoHistories()
+% Execute only demo plotting with histories of unequal lengths in any order.
+root=fileparts(fileparts(mfilename('fullpath')));
+source=fileread(fullfile(root,'Demos','d10_QualityMeasurements.m'));
+first=strfind(source,'%%RMSE plot');
+last=strfind(source,'%% Compute the parameters for the target image');
+plotting=source(first(1):last(1)-1);
+previousVisibility=get(groot,'DefaultFigureVisible');
+cleanup=onCleanup(@() restoreFigures(previousVisibility));
+set(groot,'DefaultFigureVisible','off');
+names={'SIRT','SART','OS-SART'};
+metrics={'RMSE','CC','MSSIM','UQI'};
+for lengths={[8 3 11],[11 8 3],[3 11 8],[0 3 1]}
+    close('all');
+    n=lengths{1};
+    qualitySIRT=reshape(1:4*n(1),4,n(1));
+    qualitySART=reshape(1:4*n(2),4,n(2))+100;
+    qualityOSSART=reshape(1:4*n(3),4,n(3))+200;
+    histories={qualitySIRT,qualitySART,qualityOSSART};
+    eval(plotting);
+    figures=findall(groot,'Type','figure');
+    assert(numel(figures)==4,'Expected one plot per metric');
+    for figureHandle=figures'
+        ax=findobj(figureHandle,'Type','axes');
+        row=find(cellfun(@(name) ~isempty(strfind(ax.Title.String,name)),metrics));
+        assert(isscalar(row),'Expected a unique metric title');
+        lines=findobj(ax,'Type','line');
+            assert(numel(lines)==nnz(n),'Expected one line per nonempty history');
+        for line=lines'
+            index=find(strcmp(line.DisplayName,names));
+            assert(isscalar(index),'Expected the correct algorithm label');
+            assert(isequal(line.XData,1:n(index)));
+            assert(isequal(line.YData,histories{index}(row,:)));
+        end
+    end
+end
+fprintf('PASS quality demo: unequal and empty histories retain metric rows and iteration counts\n');
+end
+
+function restoreFigures(visibility)
+close('all');
+set(groot,'DefaultFigureVisible',visibility);
+end