Circular Bar Plot in seaborn

Circular bar plots, also known as radial bar charts or circular histograms, are a visually appealing way to display data. In this article, we’ll explore how to create circular bar plots using the Seaborn library in Python.

What are circular bar plots?

In contrast to conventional bar charts, which display the bars horizontally or vertically, circular bar plots, sometimes called radial bar charts or circular histograms, display the bars radially around a circle. Every bar in the plot denotes a category, and the value of each category is represented by the bar’s length. The figure exhibits radial symmetry due to the bars’ uniform spacing around the circle.

Benefits of Circular Bar Plots:

• Cyclic Data Visualization: Hourly patterns, monthly trends, or directional data are examples of cyclic or periodic data that work well when displayed using circular bar charts.
• Space Efficiency: They utilize space well, particularly when handling a large number of categories.

Steps to Build a Circular Bar Plot Using Seaborn

Importing Necessary Libraries

Python3

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

Creating Sample Data

Python3

categories = ['A', 'B', 'C', 'D', 'E']
values = [20, 35, 30, 25, 40]

# Create a DataFrame from the data
data = pd.DataFrame({'categories': categories, 'values': values})

Plotting Circular bar Plot

1. Convert Values to Radians: The np.deg2rad function is used to convert the values in the ‘values’ column of the DataFrame to radians. This is necessary because the radial axis in the circular bar plot represents angles, and values need to be converted to radians to correctly position the bars around the circle.
2. Create a Figure and Axis: The plt.subplots function is used to create a figure (fig) and a set of subplots (ax) with a polar projection (subplot_kw=dict(polar=True)). This means that the plot will be circular, with the radial axis representing angles.
3. Plot the Bars: The sns.barplot function is used to plot the bars on the circular plot. The ‘categories’ column is used as the x-axis variable (x='categories'), the ‘angle’ column (converted to radians) is used as the y-axis variable (y='angle'), and the ‘categories’ column is also assigned to the hue parameter to color the bars differently for each category. The palette='viridis' argument specifies the color palette to use.
4. Set the Radial Axis Label: The ax.set_rticks([]) function is used to remove the radial axis labels. Since the radial axis represents angles and not numerical values, the labels are unnecessary in this context.

Python3

# Convert values to radians
data['angle'] = np.deg2rad(data['values'])

# Create a figure and axis
fig, ax = plt.subplots(figsize=(6, 6), subplot_kw=dict(polar=True))

# Plot the bars
sns.barplot(x='categories', y='angle', hue='categories',
            data=data, ax=ax, palette='viridis', legend=False)

# Set the radial axis label
ax.set_rticks([])

# Display the plot
plt.show()

Output:

Circular Bar Plot

Adding Labels, Titles

• Labeling Categories on the Plot: We can add text labels to each bar to indicate the category it represents and the value it corresponds to.
• Adding a Title to the Plot: We can add a descriptive title to the plot to provide context and make it easier for viewers to understand the purpose of the visualization.

The ax.set_title('Circular Bar Plot', pad=20) function is used to add a title to the plot, with a padding of 20 units to adjust the distance between the title and the plot.

Python3

# Convert values to radians
data['angle'] = np.deg2rad(data['values'])

# Create a figure and axis
fig, ax = plt.subplots(figsize=(6, 6), subplot_kw=dict(polar=True))

# Plot the bars
bars = sns.barplot(x='categories', y='angle', hue='categories', data=data,
                   ax=ax, palette='viridis', edgecolor='none', alpha=0.8, linewidth=0)

# Set the radial axis label
ax.set_rticks([])

# Add text labels to the bars
for bar, value in zip(bars.patches, data['values']):
    ax.text(bar.get_x() + bar.get_width() / 2,
            bar.get_height() + 0.1,
            str(value),
            ha='center', va='bottom', color='black', fontsize=10)

# Add category labels
ax.set_xticks(np.linspace(0, 2*np.pi, len(categories), endpoint=False))
ax.set_xticklabels(categories)

# Add a title
plt.title('Circular Bar Plot', pad=20)

# Display the plot
plt.show()

Output:

Adding titles

Reordering Circular Bar Plot

To reorder the categories in the circular bar plot, we can customize the order of categories in the DataFrame before plotting.

Python3

# Create a DataFrame from the data
data = pd.DataFrame({'categories': categories, 'values': values})

# Define the desired order of categories
custom_order = ['B', 'C', 'A', 'E', 'D']

# Reorder the DataFrame based on the custom order
data['categories'] = pd.Categorical(data['categories'],
                                    categories=custom_order, ordered=True)
data = data.sort_values('categories')

# Convert values to radians
data['angle'] = np.deg2rad(data['values'])

# Create a figure and axis
fig, ax = plt.subplots(figsize=(6, 6), subplot_kw=dict(polar=True))

# Plot the bars
bars = sns.barplot(x='categories', y='angle', hue='categories',
                   data=data, ax=ax, palette='viridis')

# Set the radial axis label
ax.set_rticks([])

# Add text labels to the bars
for bar, value in zip(bars.patches, data['values']):
    ax.text(bar.get_x() + bar.get_width() / 2,
            bar.get_height() + 0.1,
            str(value),
            ha='center', va='bottom')

# Add category labels
ax.set_xticks(np.linspace(0, 2*np.pi, len(custom_order),
                          endpoint=False))
ax.set_xticklabels(custom_order)

# Add a title
ax.set_title('Circular Bar Plot', pad=20)

# Add legend
handles, labels = ax.get_legend_handles_labels()
ax.legend(handles, labels, title='Categories', loc='upper right',
          bbox_to_anchor=(1.3, 1))

# Display the plot
plt.show()

Output:

Reordered Plot

Circular Bar Plot for grouped data

To create a grouped circular bar plot, we can use the hue parameter in the sns.barplot() function to differentiate between different groups of bars.

Python3

# Sample data
categories = ['A', 'B', 'C', 'D', 'E']
values1 = [20, 35, 30, 25, 40]
values2 = [15, 25, 20, 30, 35]
groups = ['Group 1', 'Group 2', 'Group 1', 'Group 2', 'Group 1']

# Create a DataFrame from the data
data = pd.DataFrame({'categories': categories, 'values1': values1,
                     'values2': values2, 'groups': groups})

# Convert values to radians
data['angle1'] = np.deg2rad(data['values1'])
data['angle2'] = np.deg2rad(data['values2'])

# Create a figure and axis
fig, ax = plt.subplots(figsize=(8, 8), subplot_kw=dict(polar=True))

# Plot the first group of bars
sns.barplot(x='categories', y='angle1', hue='groups', data=data,
            ax=ax, palette='Set1')

# Plot the second group of bars, offsetting them slightly to avoid overlap
sns.barplot(x='categories', y='angle2', hue='groups', data=data, ax=ax,
            palette='Set2', alpha=0.7)

# Set the radial axis label
ax.set_rticks([])

# Add a title
ax.set_title('Grouped Circular Bar Plot', pad=20)

# Add legend
handles, labels = ax.get_legend_handles_labels()
ax.legend(handles, labels, title='Groups', loc='upper right', bbox_to_anchor=(1.3, 1))

# Display the plot
plt.show()

Output:

Circular Bar Plot For grouped dat