Data Visualization using Seaborn
Seaborn is a Python data visualization library built on top of Matplotlib that provides a high-level interface for drawing attractive and informative statistical graphics. Below is a comprehensive guide covering various types of Seaborn plots with explanations, syntax, and example code.
Table of Contents
- Relational Plots
- Distribution Plots
- Categorical Plots
- Regression Plots
- Matrix Plots
- Multi-plot Grids
- Styling and Themes
Relational Plots
Used to visualize relationships between variables.
Scatter Plot (relplot or scatterplot)
Syntax:
sns.scatterplot(x, y, data, hue, size, style)
# or
sns.relplot(x, y, data, kind='scatter', hue, size, style)
Example:
import seaborn as sns
import matplotlib.pyplot as plt
tips = sns.load_dataset('tips')
# Using scatterplot
plt.figure(figsize=(8, 6))
sns.scatterplot(x='total_bill', y='tip', data=tips, hue='time', size='size')
plt.title('Scatter Plot of Tips vs Total Bill')
plt.show()
# Using relplot
sns.relplot(x='total_bill', y='tip', data=tips, kind='scatter',
hue='day', style='time', size='size')
plt.show()
Line Plot (lineplot)
Syntax:
sns.lineplot(x, y, data, hue, style, ci)
Example:
fmri = sns.load_dataset('fmri')
plt.figure(figsize=(10, 6))
sns.lineplot(x='timepoint', y='signal', data=fmri,
hue='region', style='event', ci=68)
plt.title('FMRI Signal Over Time')
plt.show()
Distribution Plots
Used to visualize distributions of data.
Histogram (histplot)
Syntax:
sns.histplot(data, x, y, bins, kde, hue)
Example:
penguins = sns.load_dataset('penguins')
plt.figure(figsize=(8, 6))
sns.histplot(data=penguins, x='flipper_length_mm',
bins=20, kde=True, hue='species')
plt.title('Distribution of Flipper Length')
plt.show()
Kernel Density Estimate (kdeplot)
Syntax:
sns.kdeplot(data, x, y, hue, shade)
Example:
plt.figure(figsize=(8, 6))
sns.kdeplot(data=penguins, x='bill_length_mm',
hue='species', shade=True)
plt.title('KDE of Bill Length')
plt.show()
Joint Plot (jointplot)
Syntax:
sns.jointplot(x, y, data, kind='scatter'|'kde'|'hist'|'reg'|'hex')
Example:
sns.jointplot(x='total_bill', y='tip', data=tips,
kind='reg', hue='time')
plt.show()
Pair Plot (pairplot)
Syntax:
sns.pairplot(data, hue, vars, kind, diag_kind)
Example:
sns.pairplot(penguins, hue='species',
vars=['bill_length_mm', 'bill_depth_mm', 'flipper_length_mm'])
plt.show()
Categorical Plots
Used to visualize categorical data.
Bar Plot (barplot)
Syntax:
sns.barplot(x, y, data, hue, ci, estimator)
Example:
plt.figure(figsize=(8, 6))
sns.barplot(x='day', y='total_bill', data=tips,
hue='sex', ci=68, estimator=sum)
plt.title('Total Bill by Day and Gender')
plt.show()
Count Plot (countplot)
Syntax:
sns.countplot(x, data, hue)
Example:
plt.figure(figsize=(8, 6))
sns.countplot(x='day', data=tips, hue='sex')
plt.title('Count of Customers by Day')
plt.show()
Box Plot (boxplot)
Syntax:
sns.boxplot(x, y, data, hue)
Example:
plt.figure(figsize=(10, 6))
sns.boxplot(x='day', y='total_bill', data=tips, hue='time')
plt.title('Box Plot of Total Bill by Day and Time')
plt.show()
Violin Plot (violinplot)
Syntax:
sns.violinplot(x, y, data, hue, split)
Example:
plt.figure(figsize=(10, 6))
sns.violinplot(x='day', y='total_bill', data=tips,
hue='sex', split=True)
plt.title('Violin Plot of Total Bill by Day and Gender')
plt.show()
Swarm Plot (swarmplot)
Syntax:
sns.swarmplot(x, y, data, hue)
Example:
plt.figure(figsize=(10, 6))
sns.swarmplot(x='day', y='total_bill', data=tips, hue='sex')
plt.title('Swarm Plot of Total Bill by Day')
plt.show()
Regression Plots
Used to visualize relationships with regression fits.
Regression Plot (regplot)
Syntax:
sns.regplot(x, y, data, ci, order)
Example:
plt.figure(figsize=(8, 6))
sns.regplot(x='total_bill', y='tip', data=tips,
ci=95, order=2)
plt.title('Quadratic Regression of Tips vs Total Bill')
plt.show()
LM Plot (lmplot)
Syntax:
sns.lmplot(x, y, data, hue, col, row)
Example:
sns.lmplot(x='total_bill', y='tip', data=tips,
hue='smoker', col='time', row='sex')
plt.show()
Matrix Plots
Used for matrix-like data visualization.
Heatmap (heatmap)
Syntax:
sns.heatmap(data, annot, fmt, cmap)
Example:
flights = sns.load_dataset('flights')
flights_matrix = flights.pivot('month', 'year', 'passengers')
plt.figure(figsize=(10, 8))
sns.heatmap(flights_matrix, annot=True, fmt='d', cmap='YlGnBu')
plt.title('Passengers Heatmap by Year and Month')
plt.show()
Cluster Map (clustermap)
Syntax:
sns.clustermap(data, standard_scale, z_score)
Example:
plt.figure(figsize=(10, 8))
sns.clustermap(flights_matrix, standard_scale=1, cmap='coolwarm')
plt.show()
Multi-plot Grids
For creating multiple plots in a grid.
Facet Grid (FacetGrid)
Syntax:
g = sns.FacetGrid(data, col, row, hue)
g.map(plot_type, x, y)
Example:
g = sns.FacetGrid(tips, col='time', row='smoker', hue='sex')
g.map(sns.scatterplot, 'total_bill', 'tip')
g.add_legend()
plt.show()
Pair Grid (PairGrid)
Syntax:
g = sns.PairGrid(data, vars, hue)
g.map_upper(func)
g.map_diag(func)
g.map_lower(func)
Example:
g = sns.PairGrid(penguins, vars=['bill_length_mm', 'bill_depth_mm', 'flipper_length_mm'], hue='species')
g.map_upper(sns.scatterplot)
g.map_diag(sns.histplot)
g.map_lower(sns.kdeplot)
g.add_legend()
plt.show()
Styling and Themes
Seaborn provides several built-in themes.
Syntax:
sns.set_style('darkgrid'|'whitegrid'|'dark'|'white'|'ticks')
sns.set_context('paper'|'notebook'|'talk'|'poster')
sns.set_palette('palette_name')
Example:
sns.set_style('whitegrid')
sns.set_palette('husl')
sns.set_context('talk')
plt.figure(figsize=(8, 6))
sns.barplot(x='day', y='total_bill', data=tips)
plt.title('Styled Bar Plot')
plt.show()
Common Seaborn Functions
| Function | Description |
|---|---|
sns.load_dataset() |
Load example datasets |
sns.set() |
Set aesthetic parameters |
sns.despine() |
Remove spines from plot |
sns.color_palette() |
Return a list of colors |
sns.palplot() |
Plot a color palette |
sns.axes_style() |
Return the parameters of the style |
Seaborn works best with Pandas DataFrames, where variables are stored in columns. Most functions accept x, y, and hue parameters to specify variables from the DataFrame.