Anaconda Navigator中Jupyter Notebook无法导入已安装的nglview求助

问题背景 #

在Anaconda Navigator GUI的base（根）环境启动的Jupyter Notebook中，无法导入已安装的nglview，相关版本信息：

• Jupyter Notebook 6.4.12
• Anaconda Navigator 2.3.2
• Python 3.9
• nglview 3.0.3
• ipython 8.5.0

导入时触发如下错误：

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
Cell In [1], line 5
      3 import pandas as pd
      4 import matplotlib.pyplot as plt
----> 5 import nglview as nv
      7 # the next line is necessary to display plots in Jupyter
      8 get_ipython().run_line_magic('matplotlib', 'inline')

File ~\anaconda3\lib\site-packages\nglview\__init__.py:4
      1 import warnings
      3 # for doc
----> 4 from . import adaptor, datafiles, show, widget
      5 from ._version import get_versions
      6 from .adaptor import *

File ~\anaconda3\lib\site-packages\nglview\show.py:13
      3 from . import datafiles
      4 from .adaptor import (ASEStructure, ASETrajectory, BiopythonStructure,
      5                       FileStructure, HTMDTrajectory, IODataStructure,
      6                       IOTBXStructure, MDAnalysisTrajectory, MDTrajTrajectory,
   (...)\n     11                       RdkitStructure,
     12                       TextStructure)
---> 13 from .widget import NGLWidget
     15 __all__ = [
     16     'demo',
     17     'show_pdbid',
   (...)\n     40     'show_biopython',
     41 ]
     44 def show_pdbid(pdbid, **kwargs):

File ~\anaconda3\lib\site-packages\nglview\widget.py:19
     15 from traitlets import (Bool, CaselessStrEnum, Dict, Instance, Int, Integer,
     16                        List, Unicode, observe, validate)
     17 import traitlets
---> 19 from . import color, interpolate
     20 from .adaptor import Structure, Trajectory
     21 from .component import ComponentViewer

File ~\anaconda3\lib\site-packages\nglview\color.py:114
    110         else:
    111             raise ValueError(f"{obj} must be either list of list or string")
--> 114 ColormakerRegistry = _ColormakerRegistry()

File ~\anaconda3\lib\site-packages\nglview\base.py:10, in _singleton.<locals>.getinstance()
      8 def getinstance():
      9     if cls not in instances:
---> 10         instances[cls] = cls()
     11     return instances[cls]

File ~\anaconda3\lib\site-packages\nglview\color.py:47, in _ColormakerRegistry.__init__(self, *args, **kwargs)
     45 try:
     46     get_ipython() # only display in notebook
---> 47     self._ipython_display_()
     48 except NameError:
     49     pass

File ~\anaconda3\lib\site-packages\nglview\color.py:54, in _ColormakerRegistry._ipython_display_(self, **kwargs)
     52 if self._ready:
     53     return
---> 54 super()._ipython_display_(**kwargs)

AttributeError: 'super' object has no attribute '_ipython_display_'

已尝试无效操作：

• 重启Anaconda Navigator
• 重启Jupyter服务器
• 在Notebook单元格执行%conda install -c conda-forge nglview（提示已安装）
• 关闭相关标签页

限制条件：

• 仅允许通过Anaconda Navigator GUI操作（面向无计算科学背景学生）
• 后台有无法中断的量子化学计算，不能重启电脑

解决方案（仅GUI操作） #

方案1：降级nglview到兼容版本 #

• 打开Anaconda Navigator，切换到Home标签页，确认当前环境为base (root)
• 找到nglview，点击右侧的齿轮图标，选择Mark for specific version installation
• 在版本列表中选择2.7.7（该版本与Python 3.9、IPython 8.5.0兼容性更稳定）
• 点击页面底部的Apply按钮，等待安装完成
• 重启Jupyter Notebook，尝试导入nglview

方案2：创建独立新环境 #

• 打开Anaconda Navigator，切换到Environments标签页
• 点击页面底部的Create按钮，输入环境名称（如nglview_env），选择Python版本为3.9
• 点击Create等待环境创建完成
• 在新环境的搜索框输入nglview，勾选对应包，同时勾选jupyter notebook
• 点击Apply安装所有选中的包
• 切换到Home标签页，在新环境下启动Jupyter Notebook，尝试导入nglview

方案3：更新ipywidgets到兼容版本 #

• 在Anaconda Navigator的Home标签页，base环境下找到ipywidgets
• 点击齿轮图标，选择Mark for update（或直接选择最新兼容版本）
• 点击Apply完成更新
• 重启Jupyter Notebook后测试导入

内容的提问来源于stack exchange，提问作者Anthony Nash