Displaying Matplotlib Plots Inline in Jupyter Notebook

Jupyter Notebook is a powerful interactive computing environment, and Matplotlib is a fundamental Python library for creating visualizations. A common issue new users face is getting Matplotlib plots to display within the notebook itself, rather than as separate figure windows or text-based representations. This tutorial will guide you through the necessary steps to ensure your Matplotlib plots are displayed inline in Jupyter Notebook.

Understanding the Problem

By default, Matplotlib might not display plots directly within the notebook. Instead, you might see a textual representation like <matplotlib.figure.Figure at 0x...>, or a separate figure window may pop up. This happens because Jupyter Notebook needs to be configured to interpret Matplotlib’s plotting commands correctly. The key is to set the appropriate "backend" for Matplotlib.

The %matplotlib inline Magic Command

The simplest way to display plots inline is to use a "magic command" provided by IPython (the kernel that powers Jupyter Notebook). Place the following command in a code cell at the very beginning of your notebook:

%matplotlib inline

This command tells Matplotlib to use the inline backend, which renders the plot as a static image within the notebook’s output. After executing this cell, any subsequent plotting commands will display the results inline.

Example

Here’s a complete example:

%matplotlib inline

import matplotlib.pyplot as plt
import numpy as np

x = np.linspace(0, 3 * np.pi, 500)
plt.plot(x, np.sin(x**2))
plt.title('A simple chirp')
plt.show()

Executing this code cell will now render the plot directly within the notebook’s output.

Alternative Backends for Interactivity

While %matplotlib inline is the most common approach, there are alternative backends that offer more interactivity:

• %matplotlib notebook: This backend provides an interactive plot that allows you to zoom, pan, and rotate the plot directly within the notebook. It requires a more recent version of Matplotlib and IPython.
• %matplotlib nbagg: Similar to notebook, nbagg also provides interactive plots, but might be more compatible with older versions of Matplotlib.

Choose the backend that best suits your needs and the versions of the libraries you’re using.

Making the Configuration Permanent

If you want to avoid having to type %matplotlib inline at the beginning of every notebook, you can configure it as the default backend.

1. Locate the Configuration File: The configuration file is typically located at ~/.ipython/profile_default/ipython_config.py. If the file does not exist, you may need to create it.

2. Add the Configuration Line: Open the ipython_config.py file in a text editor and add the following line:

   c.InteractiveShellApp.matplotlib = 'inline'
   

   Save the file.

Now, whenever you start a new Jupyter Notebook, Matplotlib will automatically use the inline backend.

Troubleshooting

• Ensure the Magic Command is in a Code Cell: The %matplotlib inline command must be executed in a code cell, not a Markdown cell.
• Restart the Kernel: If you’ve made changes to the configuration file or are still having problems, try restarting the Jupyter Notebook kernel.
• Check Matplotlib and IPython Versions: Ensure you have reasonably up-to-date versions of Matplotlib and IPython. Older versions might not support all the backends.
• Clear Output: Sometimes, clearing the output of the cell containing %matplotlib inline and re-running it can resolve display issues.