def parser_one_line()

in recommenders/models/newsrec/io/mind_all_iterator.py [0:0]

• 73 lines of code
• 6 McCabe index (conditional complexity)

    def parser_one_line(self, line):
        """Parse one string line into feature values.

        Args:
            line (str): a string indicating one instance.

        Yields:
            list: Parsed results including label, impression id , user id,
            candidate_title_index, clicked_title_index,
            candidate_ab_index, clicked_ab_index,
            candidate_vert_index, clicked_vert_index,
            candidate_subvert_index, clicked_subvert_index,
        """
        if self.npratio > 0:
            impr_label = self.labels[line]
            impr = self.imprs[line]

            poss = []
            negs = []

            for news, click in zip(impr, impr_label):
                if click == 1:
                    poss.append(news)
                else:
                    negs.append(news)

            for p in poss:
                candidate_title_index = []
                impr_index = []
                user_index = []
                label = [1] + [0] * self.npratio

                n = newsample(negs, self.npratio)
                candidate_title_index = self.news_title_index[[p] + n]
                candidate_ab_index = self.news_ab_index[[p] + n]
                candidate_vert_index = self.news_vert_index[[p] + n]
                candidate_subvert_index = self.news_subvert_index[[p] + n]
                click_title_index = self.news_title_index[self.histories[line]]
                click_ab_index = self.news_ab_index[self.histories[line]]
                click_vert_index = self.news_vert_index[self.histories[line]]
                click_subvert_index = self.news_subvert_index[self.histories[line]]
                impr_index.append(self.impr_indexes[line])
                user_index.append(self.uindexes[line])

                yield (
                    label,
                    impr_index,
                    user_index,
                    candidate_title_index,
                    candidate_ab_index,
                    candidate_vert_index,
                    candidate_subvert_index,
                    click_title_index,
                    click_ab_index,
                    click_vert_index,
                    click_subvert_index,
                )

        else:
            impr_label = self.labels[line]
            impr = self.imprs[line]

            for news, label in zip(impr, impr_label):
                candidate_title_index = []
                impr_index = []
                user_index = []
                label = [label]

                candidate_title_index.append(self.news_title_index[news])
                click_title_index = self.news_title_index[self.histories[line]]

                candidate_title_index = self.news_title_index[news]
                candidate_ab_index = self.news_ab_index[news]
                candidate_vert_index = self.news_vert_index[news]
                candidate_subvert_index = self.news_subvert_index[news]
                click_title_index = self.news_title_index[self.histories[line]]
                click_ab_index = self.news_ab_index[self.histories[line]]
                click_vert_index = self.news_vert_index[self.histories[line]]
                click_subvert_index = self.news_subvert_index[self.histories[line]]
                impr_index.append(self.impr_indexes[line])
                user_index.append(self.uindexes[line])

                yield (
                    label,
                    impr_index,
                    user_index,
                    candidate_title_index,
                    candidate_ab_index,
                    candidate_vert_index,
                    candidate_subvert_index,
                    click_title_index,
                    click_ab_index,
                    click_vert_index,
                    click_subvert_index,
                )