novelty-detection-ncm/Agnotron-model/Clustering Network/agnoPlotting_2D.m

function agnoPlotting_2D(bigY, mmNet, pcaNet, ts_idx ,iterations)

yT = bigY(:, ts_idx-1);
cT = pcaNet.bigC(:, ts_idx-1);

figure(1)
if pcaNet.learning == 1
    plot(mmNet.allErrors(1:ts_idx), 'r-');
else
    plot(mmNet.allErrors(1:ts_idx), 'k-');
end
hold on;
%refline(0, pcaNet.sigmaThresh);
plot([500.5 500.5], [0 25], 'k--')
plot([1000.5 1000.5], [0 25], 'k--')
plot(1001, mmNet.allErrors(1001), 'b.', 'MarkerSize', 12)
hold off;
title(['Total Error over Time - ts: ' num2str(ts_idx)])
xlabel('Timestep')
ylabel('Error')
xlim([995 1095])
ylim([0 10])
% pause(1e-5)
makepretty;



% activityPlot = zeros(7,2);
% threshPts = zeros(7,2);
% for cIdx = 1:7
%     activityPlot(cIdx,:) = cT(cIdx) .* (pcaNet.W(cIdx,:) ./ norm(pcaNet.W(cIdx,:)));
%     
%     effectiveExcite = 1 - exp((-pcaNet.excitability(cIdx)/5).^3);
%     actThresh = 1 - effectiveExcite; % this is now a vector; great
%     threshPts(cIdx,:) = actThresh .* (pcaNet.W(cIdx,:) ./ norm(pcaNet.W(cIdx,:)));
% end

colors = {'r.', 'y.', 'c.', 'g.', 'm.', 'k.', 'b.'};
for cIdx = 1:7
    thisClust = find(pcaNet.clusters == cIdx);
    figure(5);
    plot(bigY(1, thisClust), bigY(2, thisClust), colors{cIdx})
    %         if ts_idx == 2
    hold on;
    %         end
end
%     title('Clustered Data')

% figure(5)
% plotv(pcaNet.W(1,:)', 'r')
% %     plotv(activityPlot(1,:)', 'r*')
% %     plotv(threshPts(1,:)', 'kx')
% plotv(pcaNet.W(2,:)', 'b')
% %     plotv(activityPlot(2,:)', 'b*')
% %     plotv(threshPts(2,:)', 'kx')
% plotv(pcaNet.W(3,:)', 'g')
% %     plotv(activityPlot(3,:)', 'g*')
% %     plotv(threshPts(3,:)', 'kx')
% plotv(pcaNet.W(4,:)', 'c')
% %     plotv(activityPlot(4,:)', 'c*')
% %     plotv(threshPts(4,:)', 'kx')
% plotv(pcaNet.W(5,:)', 'm')
%     plotv(activityPlot(5,:)', 'm*')
%     plotv(threshPts(5,:)', 'kx')
%plotv(pcaNet.W(6,:)', 'k')
%     plotv(activityPlot(6,:)', 'k*')
%     plotv(threshPts(6,:)', 'kx')
%plotv(pcaNet.W(7,:)', 'y')
%     plotv(activityPlot(7,:)', 'y*')
%     plotv(threshPts(7,:)', 'kx')
hold off;
title(['Clusters and Weights, iteration: ' num2str(ts_idx)])
makepretty;



% figure(1212)
% silhouette(bigY',pcaNet.clusters)

% 
% if ts_idx == 2
%     figure(10)
%     plot(bigY(1,:), bigY(2,:), 'k.')
%     hold on;
%     plotv(pcaNet.W(1,:)', 'r')
%     plotv(pcaNet.W(2,:)', 'b')
%     plotv(pcaNet.W(3,:)', 'g')
%     plotv(pcaNet.W(4,:)', 'c')
%     plotv(pcaNet.W(5,:)', 'm')
%     plotv(pcaNet.W(6,:)', 'k')
%     plotv(pcaNet.W(7,:)', 'y')
% end

% if ts_idx == iterations
%     figure(20)
%     plot(1:100, mmNet.allErrors(1:100), 'r-');
%     hold on;
%     plot(101:200, mmNet.allErrors(101:200), 'b-');
%     plot(201:300, mmNet.allErrors(201:300), 'k-');
% end

% figure(25)
% plot(ts_idx, norm(pcaNet.W(1,:)), 'r.')
% if ts_idx == 2
%     hold on;
% end
% plot(ts_idx, norm(pcaNet.W(2,:)), 'b.')
% plot(ts_idx, norm(pcaNet.W(3,:)), 'g.')
% plot(ts_idx, norm(pcaNet.W(4,:)), 'c.')
% plot(ts_idx, norm(pcaNet.W(5,:)), 'm.')
% plot(ts_idx, norm(pcaNet.W(6,:)), 'k.')
% plot(ts_idx, norm(pcaNet.W(7,:)), 'y.')
% xlim([0 iterations])
% title('Weight Magnitudes')
% ylabel('Norm of Weight')
% xlabel('Timestep')