Seaborn in Python: Data Visualization Made Easy

Chapter 4: Categorical Data Visualization: Using Seaborn for Categorical Variables

In data analysis, a significant amount of time is spent working with categorical variables, which represent categories or groups. For example, a dataset might contain columns such as gender, product category, or geographic region. Visualizing these categorical variables effectively can provide key insights into the distribution, trends, and relationships in the data. Seaborn offers powerful tools to handle categorical data, enabling analysts and data scientists to quickly explore and communicate their findings.

In this chapter, we will dive into the various Seaborn plots that are particularly useful for categorical data visualization, such as bar plots, count plots, box plots, violin plots, strip plots, and swarm plots. Each plot serves a unique purpose and offers a different perspective on the data, allowing for better interpretation and analysis.

1. Understanding Categorical Data in Seaborn

Categorical data refers to variables that can be grouped into categories, each of which may have multiple values. These variables are often qualitative and are contrasted with continuous variables (which take numerical values). Examples of categorical data include:

• Gender (Male, Female)
• Region (North, South, East, West)
• Product Type (Electronics, Clothing, Furniture)

Seaborn makes it easy to visualize categorical data in different ways by using appropriate chart types designed for such data.

2. Bar Plot in Seaborn

A bar plot is one of the most commonly used visualizations for categorical data. It displays the distribution of a categorical variable by representing the frequency or count of each category as bars. Bar plots can also show the mean or median of a numerical variable for each category.

Code Example for Bar Plot:

import seaborn as sns

import matplotlib.pyplot as plt

# Load dataset

data = sns.load_dataset('tips')

# Create a bar plot

sns.barplot(x='day', y='total_bill', data=data)

# Customize the plot

plt.title('Average Total Bill per Day')

plt.xlabel('Day of the Week')

plt.ylabel('Average Total Bill')

plt.show()

In the above example, the barplot function calculates the average total bill for each day of the week, and the heights of the bars represent these averages.

Table for Bar Plot Example:

3. Count Plot in Seaborn

A count plot is similar to a bar plot but is used when you want to visualize the frequency or count of each category in a categorical variable. It’s particularly useful when you're dealing with categorical data that doesn't have a corresponding numerical value (i.e., the data is just about counting occurrences).

Code Example for Count Plot:

# Count plot of gender

sns.countplot(x='sex', data=data)

# Customize the plot

plt.title('Count of Male and Female')

plt.xlabel('Gender')

plt.ylabel('Count')

plt.show()

Table for Count Plot Example:

4. Box Plot in Seaborn

A box plot provides a summary of a dataset’s distribution. It shows the median, quartiles, and outliers for a numerical variable, segmented by a categorical variable. Box plots are excellent for understanding the spread and identifying any outliers in the data.

Code Example for Box Plot:

# Create a box plot

sns.boxplot(x='day', y='total_bill', data=data)

# Customize the plot

plt.title('Distribution of Total Bill by Day')

plt.xlabel('Day of the Week')

plt.ylabel('Total Bill')

plt.show()

Table for Box Plot Example:

5. Violin Plot in Seaborn

A violin plot is an enhancement of the box plot and provides more information about the distribution of data. It shows the density of the data at different values, which helps visualize the distribution of the data within each category. It combines aspects of a box plot with a kernel density plot.

Code Example for Violin Plot:

# Create a violin plot

sns.violinplot(x='day', y='total_bill', data=data)

# Customize the plot

plt.title('Violin Plot of Total Bill by Day')

plt.xlabel('Day of the Week')

plt.ylabel('Total Bill')

plt.show()

Table for Violin Plot Example:

The violin plot doesn't easily lend itself to tabular representation since it visualizes the distribution rather than summarizing specific statistics. The plot shows the shape of the distribution for total bill amounts on each day, including the presence of any bimodal distributions or heavy skewness.

6. Strip Plot in Seaborn

A strip plot is a scatter plot for categorical data where each data point is drawn along a single axis. It’s useful for visualizing the distribution of individual data points and identifying clustering patterns or overlapping values.

Code Example for Strip Plot:

# Create a strip plot

sns.stripplot(x='day', y='total_bill', data=data, jitter=True)

# Customize the plot

plt.title('Strip Plot of Total Bill by Day')

plt.xlabel('Day of the Week')

plt.ylabel('Total Bill')

plt.show()

Table for Strip Plot Example:

The strip plot is not easily represented in a table, as it visualizes individual data points. It can show how individual total bill values are distributed across different days, with potential overlap between days.

7. Swarm Plot in Seaborn

A swarm plot is similar to a strip plot but avoids overlap of data points by adjusting the positioning of each point. It is ideal for visualizing categorical data with a high density of points.

Code Example for Swarm Plot:

# Create a swarm plot

sns.swarmplot(x='day', y='total_bill', data=data)

# Customize the plot

plt.title('Swarm Plot of Total Bill by Day')

plt.xlabel('Day of the Week')

plt.ylabel('Total Bill')

plt.show()

8. Combining Multiple Categorical Plots

One of the powerful features of Seaborn is the ability to combine multiple categorical plots to get more insights from the data. For example, you can combine a box plot and a strip plot to visualize both the summary statistics and the individual data points at the same time.

Code Example for Combined Plot:

sns.boxplot(x='day', y='total_bill', data=data)

sns.stripplot(x='day', y='total_bill', data=data, color='black', alpha=0.5)

plt.title('Box Plot and Strip Plot Combined')

plt.show()