如何将分类器中信息最丰富的特征保存到变量中？(Python NLTK)

import nltk
from nltk.classify import ClassifierI

class CustomClassifier(ClassifierI):
    def __init__(self, classifier):
        self._classifier = classifier

    def most_informative_features(self, n):
        return self._classifier.most_informative_features(n)

classifier = nltk.NaiveBayesClassifier.train(train_set)
custom_classifier = CustomClassifier(classifier)

n = 10  # 指定要获取的特征数量
most_informative = custom_classifier.most_informative_features(n)