Jeffmxh_sentiment_analyse.py 文件源码

def main(path_to_data, column_to_deal, output_file, input_ncores):
    '''
    ??logging??
    '''
    logger = logging.getLogger('mylogger')  
    logger.setLevel(logging.INFO) 
    console = logging.StreamHandler()  
    console.setLevel(logging.INFO) 
    formatter = logging.Formatter('[%(levelname)-3s]%(asctime)s %(filename)s[line:%(lineno)d]:%(message)s')
    console.setFormatter(formatter)  
    logger.addHandler(console)  

    '''
    ??????
    '''
    data = pd.read_excel(path_to_data)
    logger.info("??????...")
    re_sub_vec = np.vectorize(re_sub) # ?????
    data[column_to_deal] = re_sub_vec(data[column_to_deal])
    logger.info("??????...")
    data['content_list'] = data[column_to_deal].map(sentence_split)
    seg_word = jieba4null(n_core = input_ncores)
    data.loc[:,'seg_words'] = data['content_list'].map(seg_word.cut_sentence)
    logger.info("????????...")
    pool = Pool(input_ncores)
    worker = polar_classifier()
    data['sentiment'] = pool.map(worker.multi_list_classify, data['seg_words'])
    data = data.drop(['content_list','seg_words'], axis = 1)
    logger.info("????????...")
    writer = pd.ExcelWriter(output_file)
    data.to_excel(writer, sheet_name='sheet1', encoding='utf-8', index=False)
    writer.save()
    logger.info("Task done!")