mlpPlot.m

function mlpPlot(path, learnInfo, showAfterSave)
% Ïîñòðîåíèå ãðàôèêà ïðîöåññà îáó÷åíèÿ è òåñòèðîâàíèÿ ÌÑÏ
%
% mlpPlot(path, learnInfo, showAfterSave)
%
% Arguments
% path - èìÿ ôàéëà, â êîòîðûé áóäåò ñîõðàíåí ãðàôèê
% learnInfo - èñõîäíûå äàííûå î ïðîöåññå îáó÷åíèÿ
% showAfterSave - ïîêàçûâàòü ãðàôèê ïîñëå ñîõðàíåíèÿ
%
% Example
% mlpPlot(char(path), learnInfo, true);
%
% See also
% 
%
% Revisions
% Author: Vulfin Alex, Date: 17/11/2010
% Supervisor: Vulfin Alex, Date: 17/11/2010
% Author: (Next revision author), Date: (Next revision date)

% Create figure
fig = figure(   'PaperPosition', [0.6345 6.345 20.3 15.23], ...
                'PaperSize', [20.98 29.68]);

hold on;
plot(learnInfo.training.RE, '-r');
plot(learnInfo.training.CE, ':r');
leg = {'Training Regression'}; 
leg{end + 1} = 'Training Classification';
if (~isempty(learnInfo.validation.RE))
    plot(learnInfo.validation.RE, '-g');
    plot(learnInfo.validation.CE, ':g');
    leg{end + 1} = 'Validation Regression'; 
    leg{end + 1} = 'Validation Classification';
end

if (~isempty(learnInfo.test.RE))
    plot(learnInfo.test.RE, '-b');
    plot(learnInfo.test.CE, ':b');
    leg{end + 1} = 'Test Regression'; 
    leg{end + 1} = 'Test Classification';
end
hold off

str = sprintf('Error vs. Learning Time, MSE = %15.10f', ...
    learnInfo.training.RE(end));
title(str);
xlabel('Epochs');
ylabel('MSE');
legend(leg);  

saveas(fig, path, 'png');

% åñëè ôëàã ïîêàçà ðèñóíêà ñáðîøåí, òî çàêðûòü ðèñóíîê
if(~showAfterSave)
    close(fig);
end

end % of function