def get_cluster_labels(dataname, k, X, Y, s):
trn_centroids = np.load(
'../../resources/topicreps/{}_ward_euclidean_{}.centroids.npy'.format(
dataname, k))
classes = np.array([i for i in range(len(trn_centroids))])
clf = NearestCentroid()
clf.centroids_ = trn_centroids
clf.classes_ = classes
labels = clf.predict(X)
id2i = dict()
for rid, eid in Y.items():
id2i[rid] = labels[eid]
oname = '../../resources/topicreps/{}_ward_euclidean_{}-{}.labels.pkl'.format(
dataname, k, s)
pickle.dump(id2i, open(oname, 'wb'))
print("saved to {}".format(oname))
