Pandas Pie Charts

Introduction

Pie charts are a popular way to visualize data that represents parts of a whole, showing the proportional relationship between different categories. In pandas, you can easily create pie charts using the plotting capabilities built on top of Matplotlib. This guide will walk you through creating, customizing, and interpreting pie charts with pandas to effectively visualize your data.

Pie charts work best when:

• You have a small number of categories (typically less than 7)
• You want to show proportions of a whole
• The differences between values are significant enough to be visible

Basic Pie Chart Creation

Let's start with a simple pie chart example using pandas.

Setup

First, import the necessary libraries:

python

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

# Set the style for better visualizations
plt.style.use('ggplot')

Creating a Simple Pie Chart

Let's create a basic dataset showing the distribution of expenses in a monthly budget:

python

# Create a simple Series with expense categories
expenses = pd.Series([1200, 600, 400, 300, 200], 
                     index=['Housing', 'Food', 'Transportation', 'Entertainment', 'Utilities'],
                     name='Monthly Expenses')

print(expenses)

Output:

Housing           1200
Food               600
Transportation     400
Entertainment      300
Utilities          200
Name: Monthly Expenses, dtype: int64

Now, let's visualize this data as a pie chart:

python

# Create a pie chart from the Series
expenses.plot.pie(figsize=(10, 6), autopct='%1.1f%%', startangle=90)
plt.title('Monthly Expense Distribution', fontsize=15)
plt.ylabel('')  # Hide the y-label
plt.show()

In this code:

• figsize=(10, 6): Sets the figure size
• autopct='%1.1f%%': Displays the percentage on each wedge with one decimal place
• startangle=90: Rotates the start of the pie chart by 90 degrees
• We set an empty ylabel to hide the default y-label

Customizing Pie Charts

Now, let's explore how to customize the pie chart for better visualization.

Adding Exploded Slices

You can "explode" or separate specific slices to emphasize them:

python

# Create an explode tuple - Housing and Food will be exploded
explode = (0.1, 0.1, 0, 0, 0)  # Only "explode" Housing and Food

expenses.plot.pie(figsize=(10, 6), 
                 autopct='%1.1f%%', 
                 startangle=90,
                 explode=explode,
                 shadow=True)

plt.title('Monthly Expense Distribution', fontsize=15)
plt.ylabel('')
plt.show()

Customizing Colors and Styles

Let's customize the colors and add a legend:

python

# Custom colors
colors = ['#ff9999', '#66b3ff', '#99ff99', '#ffcc99', '#c2c2f0']

expenses.plot.pie(figsize=(10, 6), 
                 autopct='%1.1f%%', 
                 startangle=90,
                 colors=colors,
                 legend=False,  # We'll add a separate legend
                 wedgeprops={'linewidth': 1, 'edgecolor': 'white'})

plt.title('Monthly Expense Distribution', fontsize=16, pad=20)
plt.ylabel('')

# Add a legend in a good position
plt.legend(expenses.index, loc='upper left', bbox_to_anchor=(1, 1))

# Tight layout to ensure everything fits
plt.tight_layout()
plt.show()

Creating Pie Charts from DataFrame Columns

Often, you'll want to create a pie chart from a column in a DataFrame. Let's see how to do this:

python

# Create a sample DataFrame
data = {
    'Category': ['Housing', 'Food', 'Transportation', 'Entertainment', 'Utilities'],
    'Amount': [1200, 600, 400, 300, 200]
}

df = pd.DataFrame(data)
print(df)

Output:

        Category  Amount
0        Housing    1200
1           Food     600
2  Transportation     400
3   Entertainment     300
4       Utilities     200

Now let's create a pie chart from the DataFrame:

python

# Create a pie chart using the DataFrame
plt.figure(figsize=(10, 6))
plt.pie(df['Amount'], labels=df['Category'], autopct='%1.1f%%', startangle=90,
        colors=colors, wedgeprops={'linewidth': 1, 'edgecolor': 'white'})
plt.title('Monthly Expense Distribution', fontsize=16)
plt.axis('equal')  # Equal aspect ratio ensures the pie chart is circular
plt.show()

Real-world Example: Product Sales Analysis

Let's use a more realistic example to analyze product sales data:

python

# Create a DataFrame with sales data
sales_data = pd.DataFrame({
    'Product': ['Laptops', 'Phones', 'Tablets', 'Smartwatches', 'Headphones', 'Accessories'],
    'Sales': [45000, 62000, 28000, 15000, 20000, 12000]
})

print(sales_data)

Output:

        Product  Sales
0       Laptops  45000
1        Phones  62000
2       Tablets  28000
3  Smartwatches  15000
4    Headphones  20000
5   Accessories  12000

Now, let's visualize this data with a pie chart:

python

# Calculate the percentage for each category
sales_data['Sales_Percent'] = sales_data['Sales'] / sales_data['Sales'].sum() * 100
sales_data = sales_data.sort_values('Sales', ascending=False)

# Highlight the top seller
explode = [0.1 if i == 0 else 0 for i in range(len(sales_data))]

# Create the pie chart
plt.figure(figsize=(12, 7))
plt.pie(sales_data['Sales'], labels=sales_data['Product'], explode=explode,
        autopct=lambda p: f'{p:.1f}%\n(${p*sum(sales_data["Sales"])/100:.0f}k)',
        startangle=90, colors=plt.cm.Paired(np.arange(len(sales_data))),
        wedgeprops={'linewidth': 1, 'edgecolor': 'white'})

plt.title('Product Sales Distribution', fontsize=16, pad=20)
plt.axis('equal')
plt.tight_layout()
plt.show()

This example showcases:

• Sorting data to improve visualization
• Highlighting the top seller with explode
• Custom labeling to show both percentage and actual value
• Using a colormap for consistent color scheme

Pie Chart Best Practices

When using pie charts, consider these best practices:

1. Limit the number of categories: Pie charts work best with 6 or fewer categories
2. Order segments logically: Sort by size (largest to smallest) or in a logical order
3. Use appropriate labels: Clear labels or a legend makes the chart more readable
4. Consider alternatives for small differences: If segments are very similar in size, a bar chart might be better
5. Use percentages: Display percentages to clearly show proportions

Alternative: Donut Chart

A donut chart is a variant of a pie chart with a hole in the center. It can be more visually appealing:

python

# Create a donut chart
plt.figure(figsize=(10, 6))
plt.pie(expenses, labels=expenses.index, autopct='%1.1f%%', startangle=90,
        colors=colors, wedgeprops={'linewidth': 1, 'edgecolor': 'white'})

# Add a circle at the center to create a donut chart
centre_circle = plt.Circle((0,0), 0.70, fc='white')
fig = plt.gcf()
fig.gca().add_artist(centre_circle)

plt.title('Monthly Expense Distribution', fontsize=16)
plt.axis('equal')
plt.tight_layout()
plt.show()

Summary

In this guide, we've learned:

• How to create basic pie charts in pandas
• Techniques for customizing pie charts (colors, exploding slices, etc.)
• Creating pie charts from both Series and DataFrames
• Best practices for using pie charts effectively
• How to create variations like donut charts

Pie charts are excellent for showing proportional data when you have a limited number of categories with significant differences. Remember that while pie charts are visually appealing, they're not always the best choice for every dataset.

Additional Resources and Exercises

Further Reading

• Pandas Visualization Documentation
• Matplotlib Pie Charts

Exercises

1. Basic Practice: Create a pie chart showing the distribution of your weekly activities (sleeping, working, leisure, etc.).

2. Intermediate Practice: Generate a pie chart of population distribution by continent using real-world data. Add appropriate styling and labels.

3. Advanced Practice: Create a dashboard with both a pie chart and a bar chart showing the same data in different ways. Compare which visualization is more effective for your specific dataset.

4. Challenge: Create a nested pie chart (a pie chart with an inner and outer ring) to show hierarchical data, such as product categories and subcategories.