close all;

% Algorithme des k-means

% Normalisation des donnees
Xn=(X-repmat(mean(X),size(X,1),1))*inv(diag(std(X,1)'));

% Classification et representation
IDX = kmeans(Xn,4);
classr(X, IDX, Pays)

% Critere de la somme des inerties
IDX2 = kmeans(Xn,4,'Start','sample','Replicates',4,'Display','iter');


% Classification par construction ascendante hierarchique

% Determination des classifications
D = pdist(Xn,'euclidean');
% Critere centroid
Z1 = linkage(D,'centroid');
figure
dendrogram(Z1,'COLORTHRESHOLD',4,'LABELS',Pays)
title('critere centroid')
T1 = cluster(Z1,'MaxClust',4);
% Critere single
Z2 = linkage(D,'single');
figure
dendrogram(Z2,'COLORTHRESHOLD',4,'LABELS',Pays)
title('critere single')
T2 = cluster(Z2,'MaxClust',4);
% % Critere average
Z3 = linkage(D,'average');
figure
dendrogram(Z3,'COLORTHRESHOLD',4,'LABELS',Pays)
title('critere average')
T3 = cluster(Z3,'MaxClust',4);
% % Critere complete
Z4 = linkage(D,'complete');
figure
dendrogram(Z4,'COLORTHRESHOLD',4,'LABELS',Pays)
title('critere complete')
T4 = cluster(Z4,'MaxClust',4);
% Critere ward
Z5 = linkage(D,'ward');
figure
dendrogram(Z5,'COLORTHRESHOLD',4,'LABELS',Pays)
title('critere ward')
T5 = cluster(Z5,'MaxClust',4);