Python - Seaborn

Seaborn in Python

Introduction

Seaborn is a powerful Python data visualization library built on top of Matplotlib. It provides a high-level interface for drawing attractive and informative statistical graphics. With its ability to work seamlessly with Pandas DataFrames and its easy-to-use functions, Seaborn makes complex plots simpler and more elegant.

This comprehensive guide covers all core functionalities of Seaborn, including:

• Introduction and installation
• Working with datasets
• Seaborn aesthetics and themes
• Distribution, categorical, and matrix plots
• Multivariate plotting
• Faceting with subplots
• Customization, color palettes, and best practices

Installation and Setup

Installing Seaborn

pip install seaborn

Importing Seaborn

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

Loading Datasets

Built-in Datasets

tips = sns.load_dataset("tips")
print(tips.head())

Using Pandas DataFrames

df = pd.read_csv('your_dataset.csv')
print(df.head())

Seaborn Aesthetics

Setting Styles

sns.set_style("darkgrid")

Available Styles

• white
• dark
• whitegrid
• darkgrid
• ticks

Setting Context

sns.set_context("notebook")  # options: paper, talk, poster

Distribution Plots

Histogram

sns.histplot(tips["total_bill"], bins=20, kde=False)
plt.title("Histogram of Total Bill")
plt.show()

Kernel Density Estimate (KDE) Plot

sns.kdeplot(tips["total_bill"], shade=True)
plt.title("KDE Plot")
plt.show()

Combined Histogram and KDE

sns.histplot(tips["total_bill"], kde=True)
plt.title("Histogram + KDE")
plt.show()

Rug Plot

sns.rugplot(tips["total_bill"])
plt.title("Rug Plot")
plt.show()

Statistical Plots

Box Plot

sns.boxplot(x="day", y="total_bill", data=tips)
plt.title("Box Plot of Total Bill by Day")
plt.show()

Violin Plot

sns.violinplot(x="day", y="total_bill", data=tips)
plt.title("Violin Plot")
plt.show()

Swarm Plot

sns.swarmplot(x="day", y="total_bill", data=tips)
plt.title("Swarm Plot")
plt.show()

Strip Plot

sns.stripplot(x="day", y="total_bill", data=tips, jitter=True)
plt.title("Strip Plot")
plt.show()

Categorical Plots

Count Plot

sns.countplot(x="day", data=tips)
plt.title("Count of Records per Day")
plt.show()

Bar Plot

sns.barplot(x="day", y="tip", data=tips)
plt.title("Bar Plot of Tips by Day")
plt.show()

Point Plot

sns.pointplot(x="time", y="total_bill", data=tips)
plt.title("Point Plot")
plt.show()

Boxen Plot (Large Datasets)

sns.boxenplot(x="day", y="total_bill", data=tips)
plt.title("Boxen Plot")
plt.show()

Matrix Plots

Heatmap

corr = tips.corr()
sns.heatmap(corr, annot=True, cmap="coolwarm")
plt.title("Correlation Heatmap")
plt.show()

Clustermap

sns.clustermap(corr, annot=True, cmap="viridis")
plt.title("Cluster Map")
plt.show()

Relational Plots

Scatter Plot

sns.scatterplot(x="total_bill", y="tip", data=tips)
plt.title("Scatter Plot of Bill vs Tip")
plt.show()

Line Plot

sns.lineplot(x="size", y="tip", data=tips)
plt.title("Line Plot of Tip by Size")
plt.show()

Relplot with hue

sns.relplot(x="total_bill", y="tip", hue="smoker", data=tips)
plt.title("Relational Plot with Hue")
plt.show()

Regression Plots

Linear Regression Plot

sns.lmplot(x="total_bill", y="tip", data=tips)

Regression with Hue

sns.lmplot(x="total_bill", y="tip", hue="smoker", data=tips)

Logistic Regression

sns.lmplot(x="total_bill", y="tip", data=tips, logistic=True)

Multivariate Plots

Pair Plot

sns.pairplot(tips)

Pair Plot with Hue

sns.pairplot(tips, hue="sex")

Joint Plot

sns.jointplot(x="total_bill", y="tip", data=tips, kind="hex")

Facet Grid and Subplots

FacetGrid with scatterplot

g = sns.FacetGrid(tips, col="sex", row="time")
g.map_dataframe(sns.scatterplot, x="total_bill", y="tip")

Catplot

sns.catplot(x="day", y="tip", hue="sex", kind="box", data=tips)

Color Palettes

Default Palette

sns.set_palette("deep")

Custom Palette

palette = sns.color_palette("husl", 4)
sns.boxplot(x="day", y="total_bill", data=tips, palette=palette)

Displaying Available Palettes

print(sns.palettes.SEABORN_PALETTES.keys())

Combining Seaborn and Matplotlib

Access Matplotlib Axes

ax = sns.boxplot(x="day", y="total_bill", data=tips)
ax.set_title("Customized Box Plot")
ax.set_xlabel("Day of Week")
ax.set_ylabel("Total Bill ($)")

Saving Figures

Saving a Plot

sns.histplot(tips["total_bill"])
plt.title("Histogram")
plt.savefig("histogram.png")

Best Practices

• Use Seaborn for quick, attractive statistical plots
• Load and clean your data using Pandas before plotting
• Use FacetGrid and catplot to visualize multiple subsets
• Always label your axes and provide titles
• Choose color palettes that are visually accessible
• Use boxenplot for large datasets
• Use jointplot and pairplot for multivariate analysis

Real-World Use Case

Visualizing Car Dataset

mpg = sns.load_dataset("mpg")

sns.scatterplot(x="horsepower", y="mpg", hue="origin", data=mpg)
plt.title("Horsepower vs MPG by Origin")
plt.show()

Box Plot by Cylinders

sns.boxplot(x="cylinders", y="mpg", data=mpg)
plt.title("MPG by Number of Cylinders")
plt.show()

Pair Plot with Hue

sns.pairplot(mpg[['mpg', 'horsepower', 'weight', 'acceleration', 'origin']], hue='origin')

Seaborn is a versatile and powerful library for creating statistical visualizations in Python. Built on top of Matplotlib, it simplifies complex plots, enhances aesthetics, and makes data storytelling more effective. Whether you're plotting distributions, relationships, or multi-dimensional datasets, Seaborn equips you with the tools to create professional-grade visualizations with minimal effort.

By mastering Seaborn's APIs and combining them with your understanding of Matplotlib, you can tailor your visualizations to communicate your data insights with clarity, beauty, and precision.