Jupyter(Python3)内核崩溃自动重启：绘图代码执行失败求助

Hey there, sorry to hear you're stuck with this kernel crash when trying to plot in Jupyter! That error message The kernel appears to have died. It will restart automatically is super frustrating, especially with such a simple plotting snippet. Let's go through some practical fixes that usually get things working again:

1. Verify & Update Matplotlib/Numpy Versions #

Outdated dependencies are a common culprit here. First, check what versions you're running:

import matplotlib
print(f"Matplotlib version: {matplotlib.__version__}")
import numpy
print(f"Numpy version: {numpy.__version__}")

If you're on a version older than Matplotlib 3.0 or Numpy 1.18, upgrade them:

• For pip environments:

  pip install --upgrade matplotlib numpy
  

• For conda environments:

  conda update matplotlib numpy
  

2. Switch Matplotlib Backends #

Matplotlib uses different "backends" to render plots, and sometimes the default one clashes with your Jupyter setup. Try these tweaks:

• First, run the magic command alone before importing libraries:

  %matplotlib inline
  

  Then run your import and plotting code separately.
• Alternatively, try the notebook backend instead:

  %matplotlib notebook
  import matplotlib.pyplot as plt
  import numpy as np
  plt.plot(np.sin(np.linspace(0,2*np.pi, 100)))
  

• If you just need to generate plots without interactive display, set a headless backend:

  import matplotlib
  matplotlib.use('Agg')
  import matplotlib.pyplot as plt
  import numpy as np
  plt.plot(np.sin(np.linspace(0,2*np.pi, 100)))
  plt.savefig('test_plot.png')  # Save instead of displaying inline
  

3. Ensure Kernel & Environment Match #

Sometimes Jupyter is using a different Python kernel than the one where you installed Matplotlib/Numpy. Check this by running:

• On Linux/macOS:

  !which python
  

• On Windows:

  !where python
  

Compare this path to the Python you used to install dependencies. If they don't match, reinstall the kernel for your correct environment:

python -m ipykernel install --user --name=your_environment_name

Then restart Jupyter and select this new kernel from the "Kernel > Change Kernel" menu.

4. Test with Minimal Code #

Strip down your code to the absolute basics to rule out any hidden issues:

%matplotlib inline
import matplotlib.pyplot as plt
plt.plot([1, 2, 3, 4])
plt.show()

If this still crashes, the problem is with your environment, not the plotting logic.

5. Check System Resources #

While your test code is tiny, sometimes low memory (RAM) can cause the kernel to crash unexpectedly. Close any unnecessary apps or browser tabs, then try running the code again.

6. Reinstall Jupyter Components #

If all else fails, a fresh install of Jupyter's kernel components might fix the issue:

# Uninstall existing components
pip uninstall -y ipykernel jupyter
# Reinstall
pip install ipykernel jupyter

Restart your Jupyter server after this and test again.

If none of these work, check the terminal where you launched Jupyter—there might be more detailed error messages there that can point to the root cause (like missing system libraries for Matplotlib).

内容的提问来源于stack exchange，提问作者R. Pellerej