Enhancing Pandas Pie Plots: Customizing Labels and Legends for Clarity

DataPlotPlus -

In this short article, you can find how to customize pandas pie plot with labels and legend.

1. Steps to customize pie plot

  • Import necessary libraries: import pandas as pd and import matplotlib.pyplot as plt.
  • Prepare your data in a pandas DataFrame.
  • Plot the pie chart using df.plot(kind='pie', ...).
  • Customize labels, explode, and other parameters as needed.
  • Add a legend using plt.legend() and display the plot with plt.show().

2. Data

Consider a dataset with cities and their corresponding values as:

import pandas as pd
import numpy as np

data = {
	'Metropolis': ['New York', 'London', 'Paris', 'Tokyo', 'Sydney', 'Berlin', 'Rome', 'Moscow', 'Toronto'],
	'Risk Level': ['Low', 'High', 'Medium', 'Low', 'Low', 'Medium', 'Low', 'Medium', 'Medium']
}

df = pd.DataFrame(data)

data looks like:

Metropolis Risk Level
0 New York Low
1 London High
2 Paris Medium
3 Tokyo Low
4 Sydney Low
5 Berlin Medium
6 Rome Low
7 Moscow Medium
8 Toronto Medium

3. Example

We are going to show the pie plot customization in 3 consecutive steps in order to illustrate the customization better.

Default pie plot

Let's see the default behavior of Pandas pie plot:

df["Risk Level"].value_counts().plot(kind="pie");

which will produce:

Customize the colors and labels

In this step we will customize the colors and the labels of the default pie chart by:

colors = ['#ff9999','#66b3ff','#99ff99','#ffcc99']
expl = [0.05] * 3
data = df["Risk Level"].value_counts().plot(kind="pie",autopct='%1.1f%%', radius=1.5, shadow=True, explode=expl, colors=colors)

which will give us:

Customize the title, legend and axis

Finally we will see how to customize legend, axis and the title of the pie chart in order to achieve:

colors = ['#ff9999','#66b3ff','#99ff99','#ffcc99']
data = df["Risk Level"].value_counts()
ax = data.plot(kind="pie", autopct='%1.1f%%', shadow=True, explode=[0.05, 0.05, 0.05], colors=colors, legend=True, title='Risk Level', ylabel='', labeldistance=None)
ax.legend(bbox_to_anchor=(1, 1.02), loc='upper left')
plt.show()

which give us the final:

Code Explanation

You can find the code explanation below:

  • Define a list of colors for the pie chart slices.
  • Calculate the counts of the "Risk Level" column
  • Plot a pie chart of the data with the following settings:
    • kind="pie" -Chart type: pie chart
    • autopct='%1.1f%%' - Display percentages with one decimal point.
    • shadow=True - Add shadow effect to the chart.
    • explode=[0.05, 0.05, 0.05] - Explode the slices slightly for emphasis. Number of explodes should match the slices to avoid error ValueError: 'explode' must be of length 'x'. You can change the values to emphasize some slices
    • Apply the predefined colors for the slices.
    • Show the
      • legend - legend=True
      • title - title='Risk Level'
      • remove ylabel - ylabel=''
    • Set label distance to None to remove labels - labeldistance=None
  • Adjust legend position to be outside the pie chart.
  • Display the pie chart.

Resources