15  Seaborn: Statistical Visualization

15.1 Welcome to Seaborn!

Seaborn is a powerful Python data visualization library built on top of Matplotlib. It makes creating beautiful, informative statistical graphics simple and intuitive. This guide will teach you the fundamentals of Seaborn for creating professional visualizations.

15.2 What is Seaborn?

Seaborn provides: - High-level interface for attractive statistical graphics - Built-in themes for professional-looking plots - Tools for visualizing distributions and relationships - Integration with pandas DataFrames - Statistical estimation and visualization

15.3 Installation and Setup

First, let’s import the libraries we’ll need:

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

# Set the default theme
sns.set_theme()

# Display plots inline
%matplotlib inline

15.4 Loading Sample Data

Seaborn comes with several built-in datasets perfect for learning:

# Load the tips dataset
tips = sns.load_dataset('tips')

# Display the first few rows
print(tips.head())
print(f"\nDataset shape: {tips.shape}")
print(f"Columns: {tips.columns.tolist()}")
   total_bill   tip     sex smoker  day    time  size
0       16.99  1.01  Female     No  Sun  Dinner     2
1       10.34  1.66    Male     No  Sun  Dinner     3
2       21.01  3.50    Male     No  Sun  Dinner     3
3       23.68  3.31    Male     No  Sun  Dinner     2
4       24.59  3.61  Female     No  Sun  Dinner     4

Dataset shape: (244, 7)
Columns: ['total_bill', 'tip', 'sex', 'smoker', 'day', 'time', 'size']

15.5 Basic Plots

1. Scatter Plot - Visualizing Relationships

Scatter plots show the relationship between two numerical variables:

# Create a simple scatter plot
plt.figure(figsize=(10, 6))
sns.scatterplot(data=tips, x='total_bill', y='tip')
plt.title('Relationship between Total Bill and Tip')
plt.show()

Adding more dimensions with color and size:

# Color by category, size by party size
plt.figure(figsize=(10, 6))
sns.scatterplot(data=tips, x='total_bill', y='tip',
                hue='time', size='size', sizes=(50, 200))
plt.title('Tips with Time of Day and Party Size')
plt.show()

15.6 Distribution Plots

3. Histogram - Data Distribution

Histograms show the frequency distribution of a variable:

plt.figure(figsize=(10, 6))
sns.histplot(data=tips, x='total_bill', bins=20, kde=True)
plt.title('Distribution of Total Bills')
plt.xlabel('Total Bill ($)')
plt.show()

4. KDE Plot - Smooth Distribution

KDE (Kernel Density Estimation) plots show smooth probability distributions:

plt.figure(figsize=(10, 6))
sns.kdeplot(data=tips, x='total_bill', hue='time', fill=True, alpha=0.5)
plt.title('Bill Distribution by Time of Day')
plt.show()

5. Box Plot - Statistical Summary

Box plots display the distribution through quartiles:

plt.figure(figsize=(10, 6))
sns.boxplot(data=tips, x='day', y='total_bill', hue='time')
plt.title('Bill Distribution by Day and Time')
plt.show()

6. Violin Plot - Distribution Shape

Violin plots combine box plots with KDE:

plt.figure(figsize=(10, 6))
sns.violinplot(data=tips, x='day', y='total_bill', hue='sex', split=True)
plt.title('Bill Distribution by Day and Gender')
plt.show()

15.7 Categorical Plots

7. Bar Plot - Category Averages

Bar plots show average values for categories:

plt.figure(figsize=(10, 6))
sns.barplot(data=tips, x='day', y='total_bill', estimator=np.mean, errorbar='sd')
plt.title('Average Bill by Day')
plt.ylabel('Average Total Bill ($)')
plt.show()

8. Count Plot - Category Frequencies

Count plots show the frequency of categorical values:

plt.figure(figsize=(10, 6))
sns.countplot(data=tips, x='day', hue='time')
plt.title('Number of Visits by Day and Time')
plt.show()

9. Strip Plot - Individual Points

Strip plots show all individual data points:

plt.figure(figsize=(10, 6))
sns.stripplot(data=tips, x='day', y='tip', hue='time', dodge=True, alpha=0.7)
plt.title('Individual Tips by Day and Time')
plt.show()

10. Swarm Plot - Non-overlapping Points

Swarm plots arrange points to avoid overlap:

plt.figure(figsize=(10, 6))
sns.swarmplot(data=tips, x='day', y='tip', hue='time', dodge=True)
plt.title('Tip Distribution (Swarm Plot)')
plt.show()

15.8 Multi-Plot Grids

11. Pair Plot - Multiple Relationships

Pair plots show relationships between all variable pairs:

# Create pair plot for numerical variables
sns.pairplot(tips, hue='time', diag_kind='kde', height=2.5)
plt.suptitle('Pair Plot of Tips Dataset', y=1.02)
plt.show()

12. FacetGrid - Subplots by Category

FacetGrid creates multiple subplots based on categories:

# Create a grid of plots by time and smoker status
g = sns.FacetGrid(tips, col='time', row='smoker', height=4, aspect=1.2)
g.map(sns.scatterplot, 'total_bill', 'tip', alpha=0.7)
g.add_legend()
plt.show()

15.9 Relationship Plots

13. Regression Plot - Linear Relationships

Regression plots fit and visualize linear relationships:

plt.figure(figsize=(10, 6))
sns.regplot(data=tips, x='total_bill', y='tip', scatter_kws={'alpha':0.5})
plt.title('Linear Relationship: Bill vs Tip')
plt.show()

14. Joint Plot - Bivariate Distribution

Joint plots combine scatter plots with distribution plots:

sns.jointplot(data=tips, x='total_bill', y='tip', hue='time', height=8)
plt.show()

15. Heatmap - Correlation Matrix

Heatmaps visualize data matrices and correlations:

# Calculate correlation matrix
numerical_cols = tips.select_dtypes(include=[np.number])
correlation = numerical_cols.corr()

# Create heatmap
plt.figure(figsize=(8, 6))
sns.heatmap(correlation, annot=True, cmap='coolwarm', center=0,
            square=True, linewidths=1, fmt='.2f')
plt.title('Correlation Heatmap')
plt.show()

15.10 Customization and Styling

Themes and Styles

Seaborn provides several built-in themes:

# Available styles: darkgrid, whitegrid, dark, white, ticks
styles = ['darkgrid', 'whitegrid', 'dark', 'white', 'ticks']

fig, axes = plt.subplots(2, 3, figsize=(15, 8))
axes = axes.flatten()

for idx, style in enumerate(styles):
    sns.set_style(style)
    plt.sca(axes[idx])
    sns.scatterplot(data=tips, x='total_bill', y='tip', alpha=0.6)
    axes[idx].set_title(f'Style: {style}')

# Remove extra subplot
axes[-1].remove()

plt.tight_layout()
plt.show()

# Reset to default
sns.set_style('darkgrid')

Color Palettes

Seaborn offers many color palettes:

# Show different color palettes
palettes = ['deep', 'muted', 'pastel', 'bright', 'dark', 'colorblind']

fig, axes = plt.subplots(2, 3, figsize=(15, 8))
axes = axes.flatten()

for idx, palette in enumerate(palettes):
    plt.sca(axes[idx])
    sns.barplot(data=tips, x='day', y='total_bill', palette=palette)
    axes[idx].set_title(f'Palette: {palette}')
    axes[idx].set_ylabel('')

plt.tight_layout()
plt.show()

Custom Colors

# Use custom color palette
custom_colors = ['#FF6B6B', '#4ECDC4', '#45B7D1', '#FFA07A']

plt.figure(figsize=(10, 6))
sns.boxplot(data=tips, x='day', y='total_bill', palette=custom_colors)
plt.title('Custom Color Palette')
plt.show()

Figure Size and Context

Control the size and scale of plot elements:

# Available contexts: paper, notebook, talk, poster
contexts = ['paper', 'notebook', 'talk', 'poster']

fig, axes = plt.subplots(2, 2, figsize=(15, 12))
axes = axes.flatten()

for idx, context in enumerate(contexts):
    sns.set_context(context)
    plt.sca(axes[idx])
    sns.scatterplot(data=tips, x='total_bill', y='tip')
    axes[idx].set_title(f'Context: {context}')

plt.tight_layout()
plt.show()

# Reset to default
sns.set_context('notebook')

15.11 Advanced Customization

# Combine multiple customizations
plt.figure(figsize=(12, 7))

# Set style and palette
sns.set_style('whitegrid')
sns.set_palette('husl')

# Create plot
ax = sns.violinplot(data=tips, x='day', y='total_bill', hue='time')

# Customize labels and title
plt.title('Restaurant Bills by Day and Time', fontsize=16, fontweight='bold')
plt.xlabel('Day of Week', fontsize=12)
plt.ylabel('Total Bill ($)', fontsize=12)

# Customize legend
plt.legend(title='Meal Time', title_fontsize=12, fontsize=10, loc='upper right')

# Add grid
plt.grid(axis='y', alpha=0.3)

plt.tight_layout()
plt.show()

15.12 Real-World Example: Complete Analysis

Let’s put it all together with a complete analysis:

# Load iris dataset
iris = sns.load_dataset('iris')

# Create a comprehensive figure
fig = plt.figure(figsize=(16, 10))

## Pair plot for overview
plt.subplot(2, 3, 1)
# Note: pair plot creates its own figure, so we'll use other plots

## Distribution of petal length
plt.subplot(2, 3, 1)
sns.histplot(data=iris, x='petal_length', hue='species', kde=True, alpha=0.6)
plt.title('Petal Length Distribution by Species')

## Relationship between petal dimensions
plt.subplot(2, 3, 2)
sns.scatterplot(data=iris, x='petal_length', y='petal_width',
                hue='species', style='species', s=100)
plt.title('Petal Dimensions by Species')

## Box plot of sepal length
plt.subplot(2, 3, 3)
sns.boxplot(data=iris, x='species', y='sepal_length', palette='Set2')
plt.title('Sepal Length by Species')
plt.xticks(rotation=45)

## Violin plot of sepal width
plt.subplot(2, 3, 4)
sns.violinplot(data=iris, x='species', y='sepal_width', palette='Set2')
plt.title('Sepal Width Distribution')
plt.xticks(rotation=45)

## Correlation heatmap
plt.subplot(2, 3, 5)
iris_numeric = iris.select_dtypes(include=[np.number])
sns.heatmap(iris_numeric.corr(), annot=True, cmap='coolwarm',
            center=0, square=True, fmt='.2f')
plt.title('Feature Correlations')

## Count plot
plt.subplot(2, 3, 6)
sns.countplot(data=iris, x='species', palette='Set2')
plt.title('Sample Count by Species')
plt.xticks(rotation=45)

plt.suptitle('Iris Dataset: Comprehensive Analysis', fontsize=16, fontweight='bold', y=1.00)
plt.tight_layout()
plt.show()

15.13 Practice Exercises

Now it’s your turn! Try these exercises:

Exercise 1: Basic Plot

Load the ‘penguins’ dataset and create a scatter plot showing the relationship between bill_length_mm and bill_depth_mm, colored by species.

# Your code here:
penguins = sns.load_dataset('penguins')
# Create your scatter plot

Exercise 2: Distribution Analysis

Using the penguins dataset, create a violin plot showing the distribution of body_mass_g for each species.

# Your code here:

Exercise 3: Multi-variable Visualization

Create a FacetGrid with the tips dataset showing scatter plots of total_bill vs tip for each combination of time (columns) and sex (rows).

# Your code here:

Exercise 4: Custom Styling

Create a box plot of your choice and customize it with: - A custom color palette - A specific style (e.g., ‘whitegrid’) - Custom title and labels - Rotated x-axis labels

# Your code here:

15.14 Solutions

Solution 1:

Click to show solution
penguins = sns.load_dataset('penguins')

plt.figure(figsize=(10, 6))
sns.scatterplot(data=penguins, x='bill_length_mm', y='bill_depth_mm',
                hue='species', style='species', s=100)
plt.title('Penguin Bill Dimensions by Species')
plt.xlabel('Bill Length (mm)')
plt.ylabel('Bill Depth (mm)')
plt.show()

Solution 2:

Click to show solution
plt.figure(figsize=(10, 6))
sns.violinplot(data=penguins, x='species', y='body_mass_g', palette='muted')
plt.title('Penguin Body Mass Distribution by Species')
plt.xlabel('Species')
plt.ylabel('Body Mass (g)')
plt.show()

Solution 3:

Click to show solution
g = sns.FacetGrid(tips, col='time', row='sex', height=4, aspect=1.2)
g.map(sns.scatterplot, 'total_bill', 'tip', alpha=0.6)
g.add_legend()
g.fig.suptitle('Tips Analysis by Time and Gender', y=1.02)
plt.show()

Solution 4:

Click to show solution
sns.set_style('whitegrid')

plt.figure(figsize=(12, 6))
custom_palette = ['#FF6B6B', '#4ECDC4', '#45B7D1']
sns.boxplot(data=penguins, x='island', y='flipper_length_mm',
            hue='species', palette=custom_palette)

plt.title('Penguin Flipper Length by Island and Species',
          fontsize=14, fontweight='bold')
plt.xlabel('Island', fontsize=12)
plt.ylabel('Flipper Length (mm)', fontsize=12)
plt.xticks(rotation=45)
plt.legend(title='Species', title_fontsize=11)
plt.tight_layout()
plt.show()

15.15 Key Takeaways

Remember:

  1. Import properly: Always import seaborn as sns and matplotlib.pyplot as plt
  2. Set themes: Use sns.set_theme() or sns.set_style() for better-looking plots
  3. Choose the right plot:
    • Scatter/Line plots: Relationships and trends
    • Histograms/KDE: Distributions
    • Box/Violin plots: Statistical summaries
    • Bar/Count plots: Categorical comparisons
    • Heatmaps: Correlations and matrices
  4. Use hue and size: Add dimensions to your plots with color and size parameters
  5. Customize: Adjust colors, labels, and styles to make plots publication-ready
  6. Multi-plots: Use FacetGrid, pair plots, and joint plots for comprehensive analysis

15.16 Common Plot Types Summary

Plot Type Use Case Key Parameters
scatterplot Relationships between variables x, y, hue, size
lineplot Trends over time/continuous data x, y, hue
histplot Distribution of single variable x, bins, kde
boxplot Statistical summary by category x, y, hue
violinplot Distribution shape by category x, y, hue, split
barplot Average values by category x, y, estimator
heatmap Matrix data and correlations data, annot, cmap
pairplot Multiple variable relationships data, hue
regplot Linear regression x, y

15.17 Next Steps

Now that you understand Seaborn basics, you’re ready to: - Explore statistical estimation functions - Learn about advanced customization with matplotlib - Create interactive visualizations with Plotly - Build dashboards with Streamlit or Dash - Work with geographical data using GeoPandas

15.18 Additional Resources

  • Seaborn Official Documentation: https://seaborn.pydata.org/
  • Seaborn Gallery: https://seaborn.pydata.org/examples/index.html
  • Color Palette Guide: https://seaborn.pydata.org/tutorial/color_palettes.html

Keep practicing and happy visualizing!