我正在尝试做一个协作推荐系统。我使用MovieLens完整的数据集https://grouplens.org/datasets/movielens/latest/。我想做一个csr矩阵，用户ID在列中，电影ID在行中，收视率作为值。有一个代码：

import pandas as pd
import numpy as np

movies = pd.read_csv('movies.csv')
ratings = pd.read_csv('ratings.csv')
movies.drop(['genres'], axis=1, inplace=True)
ratings.drop(['timestamp'], axis=1, inplace=True)

user_movie_matrix = ratings.pivot(index='movieId', columns='userId', values='rating')

我得到了这个：

Traceback (most recent call last):
  File "C:\Users\Dmitr\PycharmProjects\RecomBot\main.py", line 13, in <module>
    user_movie_matrix = ratings.pivot(index='movieId', columns='userId', values='rating')
  File "C:\Users\Dmitr\PycharmProjects\RecomBot\venv\lib\site-packages\pandas\util\_decorators.py", line 331, in wrapper
    return func(*args, **kwargs)
  File "C:\Users\Dmitr\PycharmProjects\RecomBot\venv\lib\site-packages\pandas\core\frame.py", line 8567, in pivot
    return pivot(self, index=index, columns=columns, values=values)
  File "C:\Users\Dmitr\PycharmProjects\RecomBot\venv\lib\site-packages\pandas\util\_decorators.py", line 331, in wrapper
    return func(*args, **kwargs)
  File "C:\Users\Dmitr\PycharmProjects\RecomBot\venv\lib\site-packages\pandas\core\reshape\pivot.py", line 540, in pivot
    return indexed.unstack(columns_listlike)  # type: ignore[arg-type]
  File "C:\Users\Dmitr\PycharmProjects\RecomBot\venv\lib\site-packages\pandas\core\series.py", line 4455, in unstack
    return unstack(self, level, fill_value)
  File "C:\Users\Dmitr\PycharmProjects\RecomBot\venv\lib\site-packages\pandas\core\reshape\reshape.py", line 489, in unstack
    unstacker = _Unstacker(
  File "C:\Users\Dmitr\PycharmProjects\RecomBot\venv\lib\site-packages\pandas\core\reshape\reshape.py", line 137, in __init__
    self._make_selectors()
  File "C:\Users\Dmitr\PycharmProjects\RecomBot\venv\lib\site-packages\pandas\core\reshape\reshape.py", line 185, in _make_selectors
    mask = np.zeros(np.prod(self.full_shape), dtype=bool)
ValueError: negative dimensions are not allowed

我用的是python 3.9，Pandas 1.5.3和Pycharm.
我发现这是由不同的df形状引起的，但我不明白为什么没有Nan值，以及如何修复它。