def predict(name, target):
tmp = to_sample(name, target)
svm: SVC = helper.load(name + "_model.sav")
result = svm.predict([tmp])
le: LabelEncoder = helper.load(name + "_label.sav")
result = le.inverse_transform(result)
print(result)
def predict(name, target):
stopwords = helper.read_stopwords()
features = to_sample(target, stopwords)
classifier = helper.load(name + "_model.sav")
result = classifier.predict([features])
le: LabelEncoder = helper.load(name + "_label.sav")
result = le.inverse_transform(result)
print(result)
def to_sample(name, target):
vec = helper.load(name + "_data.sav")
feature_names = vec.get_feature_names()
sample_source = [k + "=" + v for k, v in target.items()]
sample_test = np.zeros(len(feature_names))
for i in sample_source:
if i in feature_names:
sample_test[feature_names.index(i)] = 1.0
return sample_test
def get_feature(item, name="sample04"):
target = np.array([item])
union: FeatureUnion = helper.load(name + "_vec.sav")
# dvec: DictVectorizer = helper.load(name + "_dvec.sav")
# # target = dvec.transform(target)
# cvec: CountVectorizer = helper.load(name + "_cvec.sav")
# # content = cvec.transform(content)
# # TfidfTransformer().fit_transform(content)
# union = FeatureUnion(
# transformer_list=[
# ("feature", Pipeline([
# ('selector', ItemSelector(1)),
# ("dvec", dvec)
# ])),
# ("content", Pipeline([
# ('selector', ItemSelector(0)),
# ('cvec', cvec),
# ('tfidf', TfidfTransformer())
# ]))
# ],
# transformer_weights={"feature": 1.0, "content": 1.0})
return union.transform(target)
def get_feature(self, item):
target = np.array([item])
union = helper.load(self.name + "_vec.sav")
return union.transform(target)
def __init__(self, name):
self.name = name
self.model = helper.load(name + "_model.sav")
def predict(set_feature, name="sample04"):
model = helper.load(name + "_model.sav")
return model.predict(set_feature)
def load_data(name, size):
feature_list = helper.load(name + "_data.sav")
label_list = helper.load(name + "_label.sav")
print("FEATURE:{} / LABEL:{}".format(feature_list.shape, len(label_list)))
return train_test_split(feature_list, label_list, test_size=size)
