def get_keywords_and_topics()

in src/jobs/util/key_document_finder.py [0:0]

• 16 lines of code
• 5 McCabe index (conditional complexity)

    def get_keywords_and_topics(self):
        self.sorted_indices = np.argsort(-self.weighted_important_word_matrix, 1)
        num_words = self.weighted_important_word_matrix.shape[1]
        num_clusters = self.weighted_important_word_matrix.shape[0]
        self.sorted_scores = np.take_along_axis(self.weighted_important_word_matrix, self.sorted_indices, axis=1)
        self.topic_info_list = []
        self.keyword_list = []
        for cluster in range(num_clusters):
            topic_words = []
            for top_score_ref in range(num_words):
                if self.sorted_scores[cluster, top_score_ref] > 0.05:
                    topic_words.append(self.vectorizer_words[self.sorted_indices[cluster, top_score_ref]])
                if len(topic_words) > MAX_WORDS_PER_TOPIC:
                    break
            self.topic_info_list.append(", ".join(topic_words))
            self.keyword_list.append(topic_words)