I’m building a restaurant recommendation system in Python using user preferences (cuisine, price range) and cosine similarity. I’m encountering a ValueError: User preference vector and restaurant feature vectors must have the same number of dimensions. error under two scenarios: