'''
將原始資料的word特徵數字化為tfidf特徵，並將結果儲存到本地

article特徵可做類似處理

'''
import pandas as pd
from sklearn.feature_extraction.text import TfidfVectorizer
import pickle
import time

t_start = time.time()

"""=====================================================================================================================
1 資料預處理
"""
df_train = pd.read_csv('train_set.csv')
df_test = pd.read_csv('test_set.csv')

df_train.drop(columns='article', inplace=True)   #article  word_seg
df_test.drop(columns='article', inplace=True)

df_all = pd.concat(objs=[df_train, df_test], axis=0, sort=True)
y_train = (df_train['class'] - 1).values  # 演算法的分類預測結果是從0開始的，所以訓練集的分類標籤也要從0開始

"""=====================================================================================================================
2 特徵工程
"""
vectorizer = TfidfVectorizer(ngram_range=(1, 2), min_df=3, max_df=0.9, sublinear_tf=True)
vectorizer.fit(df_all['word_seg'])
x_train = vectorizer.transform(df_train['word_seg'])
x_test = vectorizer.transform(df_test['word_seg'])

"""=====================================================================================================================
3 儲存至本地
"""
data = (x_train, y_train, x_test)
with open('tfidf_word.pkl', 'wb') as f:
	pickle.dump(data, f)

t_end = time.time()
print("共耗時：{}min".format((t_end-t_start)/60))