lda.py 文件源码

def assign(self, datafile=None, outputfile=None):
        if not datafile:
            datafile = click.prompt('Data file',
                                    type=str,
                                    default='sentences_all.csv')
        datafile = self.__dest(datafile)
        self.datafile = datafile
        if not outputfile:
            datafilename, ext = os.path.splitext(datafile)
            default_outputfile = datafilename+'_result'+ext
            outputfile = click.prompt('output file',
                                      type=str,
                                      default=default_outputfile)
        assert os.path.isfile(datafile), 'No such file: %s' % datafile
        print 'start assiging'
        start = time()
        with open(datafile) as fi, open(outputfile, 'w') as fo:
            csv_reader = csv.reader(fi, delimiter=',')
            csv_writer = csv.writer(fo, delimiter=',')
            for row in csv_reader:
                out_row = row[:2]  # post_id and sentence_seq
                filtered_words = LdaUtils.filter_words(row[-1].split(' '))
                out_row.append(' '.join(map(str, self.query_tag(filtered_words))))
                csv_writer.writerow(out_row)
        print 'assigning takes: %s' % LdaUtils.human_readable_time(time() - start)