(self, words_list=None, update=True)
| 43 | |
| 44 | # init |
| 45 | def init(self, words_list=None, update=True): |
| 46 | if (~os.path.exists(self.dic_path) or ~os.path.exists(self.bow_model_path) or ~os.path.exists(self.bow_index_path) or update) and (words_list!=None): |
| 47 | word_list = self._seg_word(words_list) |
| 48 | |
| 49 | if os.path.exists(self.dic_path) and update==False: |
| 50 | with open(self.dic_path, 'rb') as f: |
| 51 | self.vectorizer = pickle.load(f) |
| 52 | else: |
| 53 | try: |
| 54 | logging.info('[Bow] start build tfidf model.') |
| 55 | if words_list==None: |
| 56 | logging.error( '[Bow] words_list is None' ) |
| 57 | self._gen_model(word_list) |
| 58 | logging.info('[Bow] build tfidf model success.') |
| 59 | except Exception as e: |
| 60 | logging.error( '[Bow] build tfidf model error，error info: {} '.format(e) ) |
| 61 | if words_list!=None: |
| 62 | self.words_list_pre = [] |
| 63 | for per_word in words_list: |
| 64 | self.words_list_pre.append( self._normalize( self._predict(per_word) )[0] ) |
| 65 | self.words_list_pre = np.array(self.words_list_pre) |
| 66 | return self |
| 67 | ''' |
| 68 | # seg word |
| 69 | def _seg_word(self, words_list, jieba_flag=True, del_stopword=False): |
nothing calls this directly
no test coverage detected