Python - Combining Matplotlib and Seaborn

Combining Matplotlib and Seaborn in Python

Python’s powerful data visualization libraries, Matplotlib and Seaborn, work seamlessly together. Matplotlib is a low-level library that offers complete control over figure elements, while Seaborn is built on top of Matplotlib and provides a high-level API for attractive and informative statistical graphics. Combining the two allows you to leverage the statistical power and aesthetic enhancements of Seaborn while customizing fine-grained layout details using Matplotlib.

This tutorial explores how to use Matplotlib and Seaborn together to create advanced, customized, and insightful visualizations. You will learn to set styles, create subplots, tweak legends, modify axes, and save figures effectively.

1. Getting Started

1.1 Installing Libraries

pip install matplotlib seaborn

1.2 Importing Required Modules

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

1.3 Creating a Sample Dataset

# Sample dataset
np.random.seed(42)
df = pd.DataFrame({
    'Category': np.random.choice(['A', 'B', 'C'], size=100),
    'Value': np.random.randn(100),
    'Group': np.random.choice(['X', 'Y'], size=100)
})
print(df.head())

2. Basic Seaborn Plot with Matplotlib Customization

2.1 Create a Seaborn Boxplot

sns.boxplot(x='Category', y='Value', data=df)
plt.title('Boxplot by Category')  # Using Matplotlib to add title
plt.xlabel('Category')           # Customize labels with Matplotlib
plt.ylabel('Value')
plt.show()

2.2 Save Figure Using Matplotlib

sns.boxplot(x='Category', y='Value', data=df)
plt.title('Boxplot Saved as PNG')
plt.savefig('boxplot.png', dpi=300)
plt.close()

3. Seaborn Styles + Matplotlib Custom Layouts

3.1 Applying Seaborn Styles Globally

sns.set_style('whitegrid')
sns.set_context('notebook')

3.2 Using Subplots from Matplotlib

fig, ax = plt.subplots(1, 2, figsize=(12, 5))

sns.boxplot(x='Category', y='Value', data=df, ax=ax[0])
ax[0].set_title('Boxplot')

sns.violinplot(x='Category', y='Value', data=df, ax=ax[1])
ax[1].set_title('Violin Plot')

plt.tight_layout()
plt.show()

4. Combining Seaborn’s Power with Matplotlib Precision

4.1 Add Horizontal Line to Seaborn Plot Using Matplotlib

ax = sns.boxplot(x='Category', y='Value', data=df)
plt.axhline(0, color='red', linestyle='--')  # Matplotlib line
ax.set_title('Boxplot with Zero Reference Line')
plt.show()

4.2 Customize Ticks, Legends, and Fonts

ax = sns.barplot(x='Category', y='Value', data=df, ci=None)
ax.set_xticklabels(['Cat A', 'Cat B', 'Cat C'], fontsize=12)
ax.set_yticks(np.arange(-2, 2.5, 0.5))
ax.set_ylabel('Score', fontsize=14)
plt.title('Custom Tick and Label Formatting', fontsize=16)
plt.show()

5. Multiple Plots with Shared Data Context

5.1 Faceted Plots Using Matplotlib Subplots

fig, axes = plt.subplots(1, 2, figsize=(12, 6))

sns.histplot(df[df['Group'] == 'X']['Value'], kde=True, ax=axes[0], color='blue')
axes[0].set_title('Group X Distribution')

sns.histplot(df[df['Group'] == 'Y']['Value'], kde=True, ax=axes[1], color='orange')
axes[1].set_title('Group Y Distribution')

plt.tight_layout()
plt.show()

6. Using Seaborn FacetGrid with Matplotlib

6.1 FacetGrid Overview

g = sns.FacetGrid(df, col='Group')
g.map_dataframe(sns.histplot, x='Value', kde=True)

# Add global Matplotlib title
plt.subplots_adjust(top=0.85)
g.fig.suptitle('Distribution by Group')
plt.show()

7. Pairing Plot Types and Custom Axes

7.1 PairGrid with Line and Scatter

g = sns.PairGrid(df, hue='Group', vars=['Value'])
g.map_diag(sns.histplot)
g.map_offdiag(sns.scatterplot)
g.add_legend()

# Add Matplotlib style edits
g.fig.suptitle('PairGrid Example', fontsize=16)
plt.subplots_adjust(top=0.9)
plt.show()

8. Advanced Aesthetics

8.1 Set Global Themes

sns.set_style('darkgrid')
sns.set_context('talk')
sns.set_palette('Set2')

sns.boxplot(x='Category', y='Value', hue='Group', data=df)
plt.title('Styled Seaborn + Matplotlib Combo')
plt.show()

8.2 Add Annotations

ax = sns.barplot(x='Category', y='Value', data=df, ci=None)
plt.title('Barplot with Annotations')

# Add annotations using Matplotlib
for i, p in enumerate(ax.patches):
    height = p.get_height()
    ax.annotate(f'{height:.2f}', (p.get_x() + p.get_width() / 2., height),
                ha='center', va='bottom', fontsize=10, color='black')

plt.show()

9. Working with Color Palettes

9.1 Set Custom Color Palette

palette = {'X': 'skyblue', 'Y': 'salmon'}
sns.boxplot(x='Category', y='Value', hue='Group', data=df, palette=palette)
plt.title('Custom Palette by Group')
plt.show()

9.2 Revert to Matplotlib Defaults

sns.reset_defaults()
plt.plot([1, 2, 3], [3, 2, 1])
plt.title('Matplotlib Default Style')
plt.show()

10. Exporting High-Quality Visuals

10.1 Save in Different Formats

sns.boxplot(x='Category', y='Value', data=df)
plt.title('Export Example')
plt.savefig('export_plot.svg', format='svg', dpi=300)
plt.close()

11. Common Mistakes and Fixes

11.1 Using plt.show() after sns.FacetGrid

g = sns.FacetGrid(df, col='Group')
g.map_dataframe(sns.histplot, x='Value')
plt.show()  # Required after .map_dataframe

11.2 Misaligned Titles or Labels

fig, ax = plt.subplots()
sns.boxplot(x='Category', y='Value', data=df, ax=ax)
ax.set_title('Correct Title Placement')  # Use ax.set_title
plt.show()

12. Use Cases and Applications

• Exploratory Data Analysis (EDA): Use Seaborn to quickly understand distributions, Matplotlib for custom annotations or layout.
• Scientific Publications: Seaborn provides visually appealing defaults; Matplotlib gives control over figure aesthetics, resolution, fonts, and export.
• Dashboards: Matplotlib helps embed Seaborn plots into GUIs or web dashboards (e.g., Dash, Streamlit).
• Animation: Matplotlib’s animation module can be used with Seaborn plots for time-series visualizations.

13. Combining with Other Libraries

13.1 With Pandas

iris = sns.load_dataset('iris')
iris.groupby('species').mean().plot(kind='bar')
plt.title('Mean Measurements per Species')
plt.show()

13.2 With NumPy Arrays

x = np.linspace(0, 10, 100)
y = np.sin(x)

plt.plot(x, y)
sns.despine()  # Remove top and right borders
plt.title('Sine Wave with Seaborn Despine')
plt.show()

Combining Seaborn and Matplotlib provides the best of both worlds: Seaborn offers elegant, simple-to-code statistical graphics, while Matplotlib provides complete control over every visual element. Whether you're building simple reports or complex analytical dashboards, the integration of these two libraries allows you to produce publication-quality graphics that are both beautiful and informative.

From customizing titles and axes to creating intricate subplot layouts and exporting high-resolution figures, mastering both libraries will enhance your data storytelling and analysis capabilities.

Use Seaborn for rapid development and aesthetics, and Matplotlib when you need precision control or integration into larger applications.