How to Make Better Looking Matplotlib Pie Charts

How to Set Up a Project Skeleton

How to Create Python Project Files with Windows 10 PowerShell 2.0+

cd ~
New-Item -ItemType "directory" -Path ".\matplotlib-bar-project"
cd matplotlib-bar-project
virtualenv venv
.\venv\Scripts\activate

To verify that the virtual environment is active, make sure (venv) is in the PowerShell command prompt. For example, (venv) PS C:\Users\username\matplotlib-bar-project>

How to Create Python Project Files with Linux Ubuntu 14.04+ or macOS

cd ~
mkdir matplotlib-bar-project
cd matplotlib-bar-project
virtualenv -p python3 venv
source venv/bin/activate

To verify that the virtual environment is active, make sure (venv) is in the terminal command prompt.

This will create the following files and folders, and activate the virtual environment.

▾ matplotlib-bar-project/
  ▸ venv/

Installing Matplotlib with Pip

This tutorial requires you to install a specific version of Matplotlib with pip3 install matplotlib==3.3.3. To get the plot to display in a window, you can install PyQt5 with pip3 install PyQt5==5.15.2.

We want to install a target version of these libraries because there may be API changes between the time this article is written to the time you read it.

How to Create a Pie Chart with Matplotlib

For our first task, we're going to generate a simple pie chart, using the plt.pie() method, and data derived from the popularity of programming languages.

import matplotlib.pyplot as plt


# Data
labels = ['C', 'Java', 'Python', 'C++', 'Other']
values = [30.5, 19.1, 17.4, 14.3, 18.7]

# This will explode 'C' by .1
offset = (0.2, 0, 0, 0, 0)

plt.pie(values,
        labels=labels,
        radius=1,
        autopct="%1.1f%%",
        shadow=True,
        explode=offset,
        counterclock=False)


plt.savefig("plot.png")
plt.show()

Explanation of the Code

First, we import the necessary libraries(numpy and matplotlib)

Lines 5-6: We declare our labels and their values

Line 9: We emphasize C++ by exploding their slices.

Lines 11-17: We create our pie instance with the plt.pie() function.

pyplot.pie() Function Parameters Explained

plt.pie() accepts a variety of parameters, but we'll only cover the most relevant ones, for now.

plt.pie(
  x,
  explode=None,  labels=None, colors=None, 
  autopct=None, shadow=False, labeldistance=1.1,
  radius=None, counterclock=True, wedgeprops=None,
  **kwargs
)

Resulting Pie Chart

The following depicts the default styling of a Matplotlib-generated pie chart. Notice the 90s drop shadow, bold colors that make text hard to read, and cramped text. We can do better.

How to Customize a Matplotlib Pie Chart

How to Make a Donut Chart with Matplotlib

The following code switches to a lighter color palette, removes the drop shadow, and draws a circular cutout at the center of the chart.

import matplotlib.pyplot as plt


# Data
labels = ['C', 'Java', 'Python', 'C++', 'Other']
values = [30.5, 19.1, 17.4, 14.3, 18.7]

# This will explode every slice
offset = (0.1, 0.1, 0.1, 0.1, 0.1)

# Add colors
colors = ['#c4f8fe', '#ffadad', '#cafec4', '#fef4c4', '#d5c4fe', '#f7fec4']
fig1, ax = plt.subplots()

patches, texts, autotexts = ax.pie(values, radius=2.2, explode=offset, labels=labels, rotatelabels=True, colors=colors, autopct='%1.1f%%', shadow=False, startangle=180, counterclock=False)

for t in texts:
    t.set_color('grey')

for at in autotexts:
    at.set_color('grey')

# Donut
hole = plt.Circle((0, 0), 1.75, fc='white')
fig = plt.gcf()
fig.gca().add_artist(hole)

# Equal aspect ratio ensures that pie is drawn as a circle
ax.axis('equal')
plt.tight_layout()

plt.savefig("plot.png")
plt.show()

How to Rotate the Labels and Values / Percentages of a Matplotlib Pie Chart

In this example, we go as far as to rotate all of the text to fit along the wedges.

import matplotlib.pyplot as plt


# Data
labels = ['C', 'Java', 'Python', 'C++', 'Other']
values = [30.5, 19.1, 17.4, 14.3, 18.7]

# This will explode every slice
offset = (0.1, 0.1, 0.1, 0.1, 0.1)

# Add colors
colors = ['#c4f8fe', '#ffadad', '#cafec4', '#fef4c4', '#d5c4fe', '#f7fec4']
fig1, ax = plt.subplots()

# pie, texts, autotexts = ax.pie(values, radius=2.2, explode=offset, labels=data, labeldistance=0.9, rotatelabels=True, colors=colors, autopct='%1.1f%%', shadow=False, startangle=180, counterclock=False)

patches, texts, autotexts = ax.pie(values, radius=2, colors=colors, startangle=180, labeldistance=0.9, labels=labels, rotatelabels=True, counterclock=False, autopct='%1.1f%%')

plt.setp(texts, rotation_mode="anchor", ha="center", va="center")
for t, at in zip(texts, autotexts):
    rot = t.get_rotation()
    t.set_rotation(rot + 90 + (1 - rot // 180) * 180)
    at.set_rotation(t.get_rotation())
    t.set_color('grey')
    at.set_color('grey')

# Donut
hole = plt.Circle((0, 0), 1.0, fc='white')
fig = plt.gcf()
fig.gca().add_artist(hole)


# Equal aspect ratio ensures that pie is drawn as a circle
ax.axis('equal')
plt.tight_layout()

plt.savefig("plot.png")
plt.show()

Conclusion

In this tutorial, we learned about Matplotlib's customization options. We started with a pie chart generated with the default styling, made changes to the wedge colors, dropshadows, font colors, and even learned how to make a donut chart.