keysearch.py 文件源码

def save(self, save_dir='./'):
        """
        Write out the built corpus to a save directory.
        """
        # Store the tag tables.
        pickle.dump((self.tagsToDocs, self.docsToTags), open(save_dir + 'tag-tables.pickle', 'wb'))

        # Store the document titles.
        pickle.dump(self.titles, open(save_dir + 'titles.pickle', 'wb'))

        # Write out the tfidf model.
        self.tfidf_model.save(save_dir + 'documents.tfidf_model')

        # Write out the tfidf corpus.
        corpora.MmCorpus.serialize(save_dir + 'documents_tfidf.mm', self.corpus_tfidf)  

        # Write out the dictionary.
        self.dictionary.save(save_dir + 'documents.dict')

        # Save the filenames.
        pickle.dump(self.files, open(save_dir + 'files.pickle', 'wb'))

        # Save the file ID and line numbers for each document.
        pickle.dump(self.doc_line_nums, open(save_dir + 'doc_line_nums.pickle', 'wb'))

        # Objects that are not saved:
        #  - stop_list - You don't need to filter stop words for new input
        #                text, they simply aren't found in the dictionary.
        #  - frequency - This preliminary word count object is only used for
        #                removing infrequent words. Final word counts are in
        #                the `dictionary` object.