def filter_mmdt_hash(name, dlt):
datas = mmdt_load(name)
print('old len: %d' % len(datas))
new_datas = list()
for data in datas:
arr_std = mmdt_std(data)
if arr_std > dlt:
new_datas.append(data)
else:
print('remove: %s' % (data))
new_datas = list(set(new_datas))
print('new len: %d' % len(new_datas))
mmdt_save(name, new_datas)
def classify(self, filename, dlt, classify_type=1):
md = self.mmdt_hash(filename)
if md:
arr_std = mmdt_std(md)
if classify_type == 1:
sim, label = self.simple_classify(md, dlt)
elif classify_type == 2:
sim, label = self.knn_classify(md, dlt)
else:
sim = 0.0
label = 'unknown'
print('%s,%f,%s,%f' % (filename, sim, label, arr_std))
else:
print('%s mmdt_hash is None' % filename)
def classify(self, filename, dlt, classify_type=1):
md = self.mmdt_hash(filename)
if md:
arr_std = mmdt_std(md)
if classify_type == 1:
data = self.simple_classify_2(md, dlt)
elif classify_type == 2:
data = self.knn_classify_2(md, dlt)
else:
data = {
"label": "unknown",
"labels": [
{
"label": "unknown",
"ratio": "100.00%",
}
],
"similars": []
}
print(data)
else:
print("%s mmdt_hash is None" % filename)
def check_mmdt_hash(self, md):
arr_std = mmdt_std(md)
if arr_std > self.mmdt_feature_dlt:
return True
return False