def feature_label(self):
'''convert text to feature and get it's label'''
data_params = {
'user2idx': self.user2idx,
'ngram2idx': self.feature2idx
}
feature_loader = FeatureLoader(**data_params)
x, y = feature_loader.load_n_gram_binary_feature_label(self.reviews)
return x, y
def load():
path = os.path.join(ku.model_root, 'gcf.pkl')
with open(path, 'rb') as f:
gc = pickle.load(f)
return gc
config = get_toy_config()
gc = GCForest(config)
training_split = int(len(reviews) * 0.8)
valid_split = training_split - int(training_split * 0.2)
reviews = sku.shuffle(reviews)
training_data = reviews[:valid_split]
valid_data = reviews[valid_split:training_split]
testing_data = reviews[training_split:]
training_x, training_y = feature_loader.load_n_gram_binary_feature_label(
training_data, ngram2idx, user2idx, sparse_tag=False)
valid_x, valid_y = feature_loader.load_n_gram_binary_feature_label(
valid_data, ngram2idx, user2idx, sparse_tag=False)
testing_x, testing_y = feature_loader.load_n_gram_binary_feature_label(
testing_data, ngram2idx, user2idx, sparse_tag=False)
gc.fit_transform(training_x, training_y, X_test=valid_x, y_test=valid_y)
acc = test(valid_x, valid_y, gc)
save(gc)
print('acc: ', acc)
def feature_label(self):
feature2idx = self.feature2idx(None)
params = {'feature2idx': feature2idx, 'user2idx': self.user2idx}
feature_loader = FeatureLoader(**params)
x, users = feature_loader.load_n_gram_binary_feature_label(self.reviews)
return x, users