def normalize_extracted_answer()

in lmms_eval/tasks/mathvista/mathvista_evals.py [0:0]

• 35 lines of code
• 9 McCabe index (conditional complexity)

    def normalize_extracted_answer(self, extraction, choices, question_type, answer_type, precision):
        """
        Normalize the extracted answer to match the answer type
        """
        if question_type == "multi_choice":
            # make sure the extraction is a string
            if isinstance(extraction, str):
                extraction = extraction.strip()
            else:
                try:
                    extraction = str(extraction)
                except:
                    extraction = ""

            # extract "A" from "(A) text"
            letter = re.findall(r"\(([a-zA-Z])\)", extraction)
            if len(letter) > 0:
                extraction = letter[0].upper()

            options = [chr(ord("A") + i) for i in range(len(choices))]

            if extraction in options:
                # convert option letter to text, e.g. "A" -> "text"
                ind = options.index(extraction)
                extraction = choices[ind]
            else:
                # select the most similar option
                extraction = self.get_most_similar(extraction, choices)
            assert extraction in choices

        elif answer_type == "integer":
            try:
                extraction = str(int(float(extraction)))
            except:
                extraction = None

        elif answer_type == "float":
            try:
                extraction = str(round(float(extraction), precision))
            except:
                extraction = None

        elif answer_type == "list":
            try:
                extraction = str(extraction)
            except:
                extraction = None

        return extraction