var_vector = TfidfVectorizer()
train_var = var_vector.fit_transform(t_df['var'])
top_100 = np.array(var_vector.get_feature_names())
tfidf_100 = np.argsort(var_vector.idf_)[::-1]
n = 100
top_n = top_100[tfidf_100][:n]
从tfidf Vectorizer中选择前100个单词后，如何将维度更新为100？