Normalizer = MinMaxScaler()
Normalizer.fit(fea)
fea = Normalizer.transform(fea)
# dist = tool.rank_dis_c(fea)
# dist = dist - np.diag(np.diag(dist))
print("------ Clustering ------")
start = time.time()
dist = cdist(fea, fea)
groupNumber = len(np.unique(labels))
K = 15 # the number of nearest neighbors for KNN graph
a = 10
# a = 10
# cl = AGDL.AGDL(dist, groupNumber, K, v)
# cluster = AGDL.AGDL(fea, dist, groupNumber, K, 5, a)
# labels_pred = np.zeros(len(labels), dtype='i')
# for i in range(len(cluster)):
# for j in range(len(cluster[i])):
# labels_pred[cluster[i][j]] = i
labels_pred = GDL.gdl(dist, groupNumber, K, a, True)
end = time.time()
print("time =", end - start)
print("------ Computing performance measure ------")
NMI = measure.NMI(labels, labels_pred)
print("NMI =", NMI)
ACC = measure.ACC(labels, labels_pred)
print("ACC =", ACC)
fea, labels = loadData.load_coil100() # K=25 u=100 v=0.1
print("data_set = COIL100 data.shape =", fea.shape)
print("------ Normalizing data ------")
# fea = tool.data_Normalized(fea)
Normalizer = MinMaxScaler()
Normalizer.fit(fea)
fea = Normalizer.transform(fea)
# u = 100
# dist = tool.rank_dis_c(fea, u)
dist = tool.rank_order_dis(fea)
group_number = len(np.unique(labels))
K = 25 # the number of nearest neighbors for KNN graph
v = 0.1
print("------ Clustering ------")
start = time.time()
labels_pred = GDL.gdl(dist, group_number, K, v, True)
end = time.time()
print("time =", end - start)
NMI = measure.NMI(labels, labels_pred)
print("NMI =", NMI)
ACC = measure.ACC(labels, labels_pred)
print("ACC =", ACC)
precision_score = measure.precision_score(labels, labels_pred)
print("precision_score =", precision_score)
