def getClustersResult():
data_name = request.args.get('dataname', '')
kmean_number = request.args.get('kmeannumber', '')
if data_name == '':
data_name = 'normOutputCleanevaluation.txt'
data_path = './dump/' + data_name
method = request.args.get('method', '')
# Use the method send to clusterize the data
t1 = time.time()
if method == 'kmean':
clusterList, barycentersList, infosList = kmeanTreatment(data_path, int(kmean_number))
elif method == 'meanshift':
clusterList, barycentersList, infosList = meanShiftTreatment(data_path)
else:
clusterList, barycentersList, infosList = kmeanTreatment(data_path, 4)
t2 = time.time()
print 'Clusterized in:', strtimedelta(t1,t2)
clusterNumber = range(len(clusterList))
cluster_mbtags = []
for i in clusterNumber:
artist_number = len(infosList[0][i])
temp_dic = {}
for k, v in infosList[2][i].iteritems():
temp_dic[k] = format(float(v)*100 / artist_number, '.2f')
cluster_mbtags.append(sorted(temp_dic.iteritems(), key=operator.itemgetter(1), reverse = True))
# Make a dump of the results!
with open('./cluster/' + method + data_name, 'wb') as f:
cPickle.dump([clusterList, barycentersList, infosList], f)
return render_template('clusteringresult.html',
datapath = data_path,
method = method,
clusterNumber = clusterNumber,
clusterList = clusterList,
barycentersList = barycentersList,
infosList = infosList,
mbtags_sorted = cluster_mbtags)
for i in feature_number:
new_vector.append(row[1][i-1])
new_data.append([new_info, new_vector])
print "New data generated"
filename = 'dataset_' + str(len(request)) + '_' + name
data_path = './dump/normOutputClean' + filename + '.txt'
# Dump the generated data
with open(data_path, 'wb') as f:
cPickle.dump(new_data, f)
print "Data dumped"
clusterList, barycentersList, infosList = kmeanTreatment(data_path, 5)
with open('./cluster/kmean5' + name + '.txt', 'wb') as f:
cPickle.dump([clusterList, barycentersList, infosList], f)
print "Kmean 5 ok and dumped"
clusterList, barycentersList, infosList = kmeanTreatment(data_path, 10)
with open('./cluster/kmean10' + name + '.txt', 'wb') as f:
cPickle.dump([clusterList, barycentersList, infosList], f)
print "Kmean 10 ok and dumped"
clusterList, barycentersList, infosList = kmeanTreatment(data_path, 20)