Seaborn_Reference

Loading views...

Seaborn Reference Guide

Complete reference for Seaborn - statistical data visualization library built on Matplotlib.

Installation and Import

Installation

pip install seaborn

Import

import seaborn as sns
import matplotlib.pyplot as plt

Convention: Import as sns. Seaborn works with Matplotlib, so import both.

Setting Styles

`sns.set_style()`

Description: Set the aesthetic style

Syntax: sns.set_style(style)

sns.set_style('darkgrid')   # Dark background with grid
sns.set_style('whitegrid')  # White background with grid
sns.set_style('dark')       # Dark background, no grid
sns.set_style('white')      # White background, no grid
sns.set_style('ticks')      # White with ticks

`sns.set_palette()`

Description: Set color palette

Syntax: sns.set_palette(palette)

sns.set_palette('deep')
sns.set_palette('muted')
sns.set_palette('pastel')
sns.set_palette('bright')
sns.set_palette('dark')
sns.set_palette('colorblind')

Distribution Plots

`sns.histplot()`

Description: Plot histogram with optional KDE

Syntax: sns.histplot(data, x, kde=False, bins='auto')

import pandas as pd

# Basic histogram
sns.histplot(data=df, x='age')

# With KDE (Kernel Density Estimate)
sns.histplot(data=df, x='age', kde=True)

# With hue (color by category)
sns.histplot(data=df, x='age', hue='gender', kde=True)
plt.show()

`sns.kdeplot()`

Description: Plot kernel density estimate

Syntax: sns.kdeplot(data, x, hue=None)

sns.kdeplot(data=df, x='age')
sns.kdeplot(data=df, x='age', hue='gender')
plt.show()

`sns.displot()`

Description: Figure-level distribution plot

Syntax: sns.displot(data, x, kind='hist', hue=None)

sns.displot(data=df, x='age', kind='hist')
sns.displot(data=df, x='age', kind='kde')
sns.displot(data=df, x='age', kind='ecdf')  # Empirical CDF
plt.show()

Categorical Plots

`sns.barplot()`

Description: Show mean values with confidence intervals

Syntax: sns.barplot(data, x, y, hue=None)

sns.barplot(data=df, x='category', y='value')
sns.barplot(data=df, x='category', y='value', hue='gender')
plt.show()

`sns.countplot()`

Description: Count occurrences of categories

Syntax: sns.countplot(data, x, hue=None)

sns.countplot(data=df, x='category')
sns.countplot(data=df, x='category', hue='gender')
plt.show()

`sns.boxplot()`

Description: Box plot showing distribution

Syntax: sns.boxplot(data, x, y, hue=None)

sns.boxplot(data=df, x='category', y='value')
sns.boxplot(data=df, x='category', y='value', hue='gender')
plt.show()

`sns.violinplot()`

Description: Violin plot (box plot + KDE)

Syntax: sns.violinplot(data, x, y, hue=None)

sns.violinplot(data=df, x='category', y='value')
sns.violinplot(data=df, x='category', y='value', hue='gender', split=True)
plt.show()

`sns.stripplot()`

Description: Scatter plot for categorical data

Syntax: sns.stripplot(data, x, y, hue=None)

sns.stripplot(data=df, x='category', y='value')
sns.stripplot(data=df, x='category', y='value', jitter=True)
plt.show()

`sns.swarmplot()`

Description: Categorical scatter with non-overlapping points

Syntax: sns.swarmplot(data, x, y, hue=None)

sns.swarmplot(data=df, x='category', y='value')
plt.show()

Relational Plots

`sns.scatterplot()`

Description: Scatter plot with optional hue/size

Syntax: sns.scatterplot(data, x, y, hue=None, size=None)

sns.scatterplot(data=df, x='x_col', y='y_col')
sns.scatterplot(data=df, x='x_col', y='y_col', hue='category')
sns.scatterplot(data=df, x='x_col', y='y_col', hue='category', size='value')
plt.show()

`sns.lineplot()`

Description: Line plot with confidence intervals

Syntax: sns.lineplot(data, x, y, hue=None)

sns.lineplot(data=df, x='time', y='value')
sns.lineplot(data=df, x='time', y='value', hue='category')
plt.show()

Matrix Plots

`sns.heatmap()`

Description: Plot rectangular data as color-encoded matrix

Syntax: sns.heatmap(data, annot=False, cmap=None)

# Correlation matrix
correlation = df.corr()
sns.heatmap(correlation, annot=True, cmap='coolwarm')
plt.show()

# Custom colormap
sns.heatmap(correlation, annot=True, cmap='viridis', fmt='.2f')
plt.show()

# With square cells
sns.heatmap(correlation, annot=True, square=True, linewidths=0.5)
plt.show()

`sns.clustermap()`

Description: Hierarchically-clustered heatmap

Syntax: sns.clustermap(data, method='average')

sns.clustermap(df.corr(), annot=True, cmap='coolwarm')
plt.show()

Regression Plots

`sns.regplot()`

Description: Plot with linear regression line

Syntax: sns.regplot(data, x, y, order=1)

sns.regplot(data=df, x='x_col', y='y_col')
plt.show()

# Polynomial regression
sns.regplot(data=df, x='x_col', y='y_col', order=2)
plt.show()

`sns.lmplot()`

Description: Figure-level regression plot

Syntax: sns.lmplot(data, x, y, hue=None)

sns.lmplot(data=df, x='x_col', y='y_col')
sns.lmplot(data=df, x='x_col', y='y_col', hue='category')
plt.show()

`sns.residplot()`

Description: Plot residuals of regression

Syntax: sns.residplot(data, x, y)

sns.residplot(data=df, x='x_col', y='y_col')
plt.show()

Multi-Plot Grids

`sns.pairplot()`

Description: Pairwise relationships in dataset

Syntax: sns.pairplot(data, hue=None)

sns.pairplot(df)
sns.pairplot(df, hue='category')
sns.pairplot(df, hue='category', diag_kind='kde')
plt.show()

`sns.FacetGrid()`

Description: Multi-plot grid for plotting conditional relationships

Syntax: sns.FacetGrid(data, col=None, row=None, hue=None)

g = sns.FacetGrid(df, col='category')
g.map(sns.histplot, 'value')
plt.show()

# With hue
g = sns.FacetGrid(df, col='category', hue='gender')
g.map(sns.scatterplot, 'x_col', 'y_col')
g.add_legend()
plt.show()

Color Palettes

`sns.color_palette()`

Description: Return or set color palette

Syntax: sns.colorpalette(palette=None, ncolors=None)

# View palette
sns.color_palette('deep')
sns.palplot(sns.color_palette('deep'))

# Sequential palettes
sns.color_palette('Blues')
sns.color_palette('Greens')

# Diverging palettes
sns.color_palette('coolwarm')
sns.color_palette('RdBu')

# Custom palette
custom = ['#FF5733', '#33FF57', '#3357FF']
sns.set_palette(custom)

Useful Functions

`sns.despine()`

Description: Remove top and right spines

Syntax: sns.despine(top=True, right=True, left=False, bottom=False)

sns.boxplot(data=df, x='category', y='value')
sns.despine()
plt.show()

`sns.set_context()`

Description: Set plotting context (scale)

Syntax: sns.set_context(context)

sns.set_context('paper')    # Smallest
sns.set_context('notebook') # Default
sns.set_context('talk')     # Larger
sns.set_context('poster')   # Largest

`sns.set()`

Description: Set multiple aesthetic parameters

Syntax: sns.set(style=None, palette=None, context=None)

sns.set(style='darkgrid', palette='muted', context='notebook')

Back to the Topic: Libraries & Tools

← Previous Next →