VS Code中.ipynb导入matplotlib遇循环导入错误，.py文件正常

• 在VS Code中运行.ipynb文件时，导入matplotlib触发循环导入错误，报错信息为：ImportError: cannot import name '_docstring' from partially initialized module 'matplotlib' (most likely due to a circular import)
• 同一Python环境下的.py文件执行相同导入及绘图代码完全正常，且工作目录中无_docstring同名文件

.ipynb文件运行代码：

import numpy as np 
import matplotlib.pyplot as plt
import cv2 as cv

完整报错栈：

ImportError                               Traceback (most recent call last)
Cell In[3], line 2
      1 import numpy as np 
----> 2 import matplotlib.pyplot as plt
      3 import cv2 as cv

File c:\Users\P****\AppData\Local\Programs\Python\Python38\lib\site-packages\matplotlib\__init__.py:113
    109 from packaging.version import parse as parse_version
    111 # cbook must import matplotlib only within function
    112 # definitions, so it is safe to import from it here.
--> 113 from . import _api, _version, cbook, _docstring, rcsetup
    114 from matplotlib.cbook import sanitize_sequence
    115 from matplotlib._api import MatplotlibDeprecationWarning

ImportError: cannot import name '_docstring' from partially initialized module 'matplotlib' (most likely due to a circular import) (c:\Users\P****\AppData\Local\Programs\Python\Python38\lib\site-packages\matplotlib\__init__.py)

同环境下可正常运行的.py代码：

import matplotlib.pyplot as plt
import numpy as np

x = np.arange(1,10,0.1)
y = x**2

plt.plot(x,y)
plt.show()

• 重启Jupyter内核：在.ipynb文件右上角点击「重启内核」按钮，清除会话缓存后重新运行代码
• 调整导入顺序：将matplotlib的导入放在所有库的最前面，避免其他库初始化时干扰matplotlib的模块加载：

  import matplotlib.pyplot as plt
  import numpy as np 
  import cv2 as cv
  

• 清除matplotlib缓存：找到Python环境中matplotlib安装目录下的__pycache__文件夹（路径示例：c:\Users\P****\AppData\Local\Programs\Python\Python38\lib\site-packages\matplotlib\__pycache__），删除该文件夹后重新运行代码
• 升级matplotlib版本：打开终端执行命令pip install --upgrade matplotlib，修复旧版本中可能存在的循环导入bug
• 更新VS Code Jupyter扩展：在VS Code扩展市场中检查Jupyter扩展是否为最新版本，更新后重启编辑器再测试

内容的提问来源于stack exchange，提问作者Prashant Tiwari