Data Visualization

This guide covers Matplotlib and Seaborn for creating beautiful and informative visualizations.

ML toolbox curriculum map (this guide)

Short map for the visualization part of the ML toolbox (all with plotting code below).

Matplotlib: line, scatter, histogram, bar & pie → Common plot types (Matplotlib), plus Seaborn variants in later sections
Underlying data prep → NumPy and Pandas guides

ML toolbox curriculum map (this guide)
Introduction
Matplotlib Basics
Figure and Axes Structure
Common Plot Types
Customizing Plots
Subplots and Multiple Plots
Seaborn Introduction
Seaborn Relational Plots
Seaborn Categorical Plots
Seaborn Distribution Plots
Seaborn Regression & Mixed Plots
Seaborn Matrix & Styling Plots
Saving Figures
Practice Exercises

Introduction

What is Matplotlib?

Matplotlib is Python's primary plotting library. It provides:

Flexibility: Create any type of plot
Control: Fine-grained control over every element
Publication quality: High-quality figures for papers

What is Seaborn?

Seaborn is built on Matplotlib and provides:

Statistical plots: Built-in statistical visualizations
Beautiful defaults: Attractive styling out of the box
Easy to use: Less code for common plots

Installation

pip install matplotlib seaborn

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

# Set style
plt.style.use('seaborn-v0_8')  # or 'default', 'ggplot', etc.
sns.set_style("whitegrid")

Matplotlib Basics

Simple Plot

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

# Create plot
plt.figure(figsize=(10, 6))
plt.plot(x, y)
plt.xlabel('X axis')
plt.ylabel('Y axis')
plt.title('Simple Sine Wave')
plt.grid(True)
plt.show()

Multiple Lines

x = np.linspace(0, 10, 100)
y1 = np.sin(x)
y2 = np.cos(x)

plt.figure(figsize=(10, 6))
plt.plot(x, y1, label='sin(x)')
plt.plot(x, y2, label='cos(x)')
plt.xlabel('X')
plt.ylabel('Y')
plt.title('Sine and Cosine')
plt.legend()
plt.grid(True)
plt.show()

Figure and Axes

# Explicit figure and axes
fig, ax = plt.subplots(figsize=(10, 6))
ax.plot(x, y)
ax.set_xlabel('X axis')
ax.set_ylabel('Y axis')
ax.set_title('Using Axes Object')
ax.grid(True)
plt.show()

Figure and Axes Structure

Understanding the structure of a Matplotlib figure is crucial for creating professional visualizations.

Components

Figure (Canvas):

The entire window or page that everything is drawn on
Can contain multiple Axes (subplots)
Controls overall size, DPI, and background

Axes (Plot Area):

The actual plot area where data is visualized
Contains the plot, labels, legend, etc.
A Figure can contain multiple Axes

Axis (x/y scales):

The number-line-like objects that provide ticks and labels
X-axis and Y-axis define the coordinate system

# Create figure and axes explicitly
fig = plt.figure(figsize=(10, 6))  # Figure (canvas)
ax = fig.add_subplot(111)          # Axes (plot area)

# Or use the convenient method
fig, ax = plt.subplots(figsize=(10, 6))

# Access axes
ax.plot(x, y)
ax.set_xlabel('X Axis')            # X-axis label
ax.set_ylabel('Y Axis')            # Y-axis label
ax.set_title('Title')              # Title on axes

# Multiple axes in one figure
fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(12, 5))
ax1.plot(x, y1)
ax2.plot(x, y2)

Understanding the Hierarchy

# Figure level
fig = plt.figure(figsize=(10, 6), dpi=100)  # DPI = dots per inch
fig.suptitle('Figure Title', fontsize=16)    # Title for entire figure

# Axes level
ax = fig.add_subplot(111)
ax.set_title('Axes Title')                   # Title for this plot
ax.set_xlabel('X Label')                     # X-axis label
ax.set_ylabel('Y Label')                     # Y-axis label

# Axis level (access through axes)
ax.xaxis.set_major_locator(plt.MaxNLocator(10))  # Control x-axis ticks
ax.yaxis.set_major_formatter(plt.FuncFormatter(lambda x, p: f'${x:.0f}'))  # Format y-axis

Common Plot Types

Line Plot

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

plt.figure(figsize=(10, 6))
plt.plot(x, y, linewidth=2, color='blue', linestyle='-', marker='o', markersize=4)
plt.xlabel('X')
plt.ylabel('Y')
plt.title('Line Plot')
plt.show()

Scatter Plot

x = np.random.randn(100)
y = np.random.randn(100)
colors = np.random.rand(100)
sizes = 1000 * np.random.rand(100)

plt.figure(figsize=(10, 6))
plt.scatter(x, y, c=colors, s=sizes, alpha=0.5, cmap='viridis')
plt.colorbar()
plt.xlabel('X')
plt.ylabel('Y')
plt.title('Scatter Plot')
plt.show()

Bar Plot

categories = ['A', 'B', 'C', 'D', 'E']
values = [23, 45, 56, 78, 32]

plt.figure(figsize=(10, 6))
plt.bar(categories, values, color='skyblue', edgecolor='black')
plt.xlabel('Categories')
plt.ylabel('Values')
plt.title('Bar Plot')
plt.show()

# Horizontal bar
plt.figure(figsize=(10, 6))
plt.barh(categories, values, color='lightcoral')
plt.xlabel('Values')
plt.ylabel('Categories')
plt.title('Horizontal Bar Plot')
plt.show()

Histogram

data = np.random.normal(100, 15, 1000)

plt.figure(figsize=(10, 6))
plt.hist(data, bins=30, color='steelblue', edgecolor='black', alpha=0.7)
plt.xlabel('Value')
plt.ylabel('Frequency')
plt.title('Histogram')
plt.show()

# Multiple histograms
data1 = np.random.normal(100, 15, 1000)
data2 = np.random.normal(120, 15, 1000)

plt.figure(figsize=(10, 6))
plt.hist(data1, bins=30, alpha=0.5, label='Group 1', color='blue')
plt.hist(data2, bins=30, alpha=0.5, label='Group 2', color='red')
plt.xlabel('Value')
plt.ylabel('Frequency')
plt.title('Multiple Histograms')
plt.legend()
plt.show()

Box Plot

data = [np.random.normal(0, std, 100) for std in range(1, 5)]

plt.figure(figsize=(10, 6))
plt.boxplot(data, labels=['Group 1', 'Group 2', 'Group 3', 'Group 4'])
plt.ylabel('Value')
plt.title('Box Plot')
plt.show()

Pie Chart

sizes = [30, 25, 25, 20]
labels = ['A', 'B', 'C', 'D']
colors = ['#ff9999', '#66b3ff', '#99ff99', '#ffcc99']

plt.figure(figsize=(8, 8))
plt.pie(sizes, labels=labels, colors=colors, autopct='%1.1f%%', startangle=90)
plt.title('Pie Chart')
plt.show()

Customizing Plots

Colors and Styles

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

plt.figure(figsize=(10, 6))
plt.plot(x, np.sin(x), color='red', linewidth=2, linestyle='--', label='sin(x)')
plt.plot(x, np.cos(x), color='#0066CC', linewidth=2, linestyle='-', marker='o', markersize=3, label='cos(x)')
plt.xlabel('X', fontsize=12, fontweight='bold')
plt.ylabel('Y', fontsize=12, fontweight='bold')
plt.title('Customized Plot', fontsize=14, fontweight='bold')
plt.legend(fontsize=10, loc='upper right')
plt.grid(True, alpha=0.3)
plt.show()

Axis Customization

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

fig, ax = plt.subplots(figsize=(10, 6))
ax.plot(x, y)
ax.set_xlabel('X axis', fontsize=12)
ax.set_ylabel('Y axis', fontsize=12)
ax.set_title('Customized Axes', fontsize=14)
ax.set_xlim(0, 10)
ax.set_ylim(-1.5, 1.5)
ax.set_xticks(np.arange(0, 11, 2))
ax.set_yticks(np.arange(-1, 2, 0.5))
ax.grid(True, linestyle='--', alpha=0.5)
plt.show()

Text and Annotations

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

plt.figure(figsize=(10, 6))
plt.plot(x, y)
plt.xlabel('X')
plt.ylabel('Y')
plt.title('Plot with Annotations')

# Add text
plt.text(5, 0.5, 'Peak', fontsize=12, ha='center')

# Add annotation with arrow
plt.annotate('Maximum', xy=(np.pi/2, 1), xytext=(4, 0.5),
             arrowprops=dict(arrowstyle='->', color='red', lw=2),
             fontsize=12)

plt.show()

Subplots and Multiple Plots

Subplots

# Create subplots
fig, axes = plt.subplots(2, 2, figsize=(12, 10))

# Plot 1
axes[0, 0].plot(np.linspace(0, 10, 100), np.sin(np.linspace(0, 10, 100)))
axes[0, 0].set_title('Sine Wave')
axes[0, 0].grid(True)

# Plot 2
axes[0, 1].scatter(np.random.randn(100), np.random.randn(100))
axes[0, 1].set_title('Scatter Plot')

# Plot 3
axes[1, 0].bar(['A', 'B', 'C', 'D'], [10, 20, 30, 40])
axes[1, 0].set_title('Bar Plot')

# Plot 4
axes[1, 1].hist(np.random.normal(0, 1, 1000), bins=30)
axes[1, 1].set_title('Histogram')

plt.tight_layout()
plt.show()

GridSpec (Advanced Layout)

from matplotlib.gridspec import GridSpec

fig = plt.figure(figsize=(12, 8))
gs = GridSpec(2, 2, figure=fig)

ax1 = fig.add_subplot(gs[0, :])  # Top row, all columns
ax2 = fig.add_subplot(gs[1, 0])  # Bottom left
ax3 = fig.add_subplot(gs[1, 1])  # Bottom right

ax1.plot(np.linspace(0, 10, 100), np.sin(np.linspace(0, 10, 100)))
ax2.scatter(np.random.randn(50), np.random.randn(50))
ax3.bar(['A', 'B', 'C'], [10, 20, 30])

plt.tight_layout()
plt.show()

Seaborn Introduction

What is Seaborn?

Seaborn is a high-level visualization library built on top of Matplotlib, designed for quick, attractive, and informative statistical plots with minimal code.

Key Features:

Statistical plots: Built-in statistical visualizations
Beautiful defaults: Attractive styling out of the box
Easy to use: Less code for common plots
Integration: Works with Pandas DataFrames

Installation and Setup

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

# Set style globally
sns.set_style("whitegrid")  # Options: darkgrid, whitegrid, dark, white, ticks
sns.set_palette("husl")     # Set color palette

Seaborn Relational Plots

Relational plots visualize relationships between two variables, helping identify trends, correlations, and distributions.

scatterplot()

Displays individual data points to show correlation:

# Load sample data
tips = sns.load_dataset('tips')

# Basic scatter plot
plt.figure(figsize=(10, 6))
sns.scatterplot(data=tips, x='total_bill', y='tip')
plt.title('Total Bill vs Tip')
plt.show()

# With hue (color by category)
sns.scatterplot(data=tips, x='total_bill', y='tip', hue='sex')
plt.show()

# With size
sns.scatterplot(data=tips, x='total_bill', y='tip', 
                hue='sex', size='size', sizes=(50, 200))
plt.show()

# With style
sns.scatterplot(data=tips, x='total_bill', y='tip', 
                hue='sex', style='time')
plt.show()

Use Cases:

Visualize relationships between numeric variables (price vs. sales)
Discover patterns in data
Identify correlations

lineplot()

Shows continuous trends or summary lines over time or categories:

# Basic line plot
sns.lineplot(data=tips, x='total_bill', y='tip')
plt.show()

# With hue
sns.lineplot(data=tips, x='total_bill', y='tip', hue='sex')
plt.show()

# With markers
sns.lineplot(data=tips, x='total_bill', y='tip', 
             hue='sex', marker='o', markersize=8)
plt.show()

# Time series example
dates = pd.date_range('2024-01-01', periods=100, freq='D')
ts_data = pd.DataFrame({
    'date': dates,
    'value': np.cumsum(np.random.randn(100))
})
sns.lineplot(data=ts_data, x='date', y='value')
plt.show()

Use Cases:

Show trends over time
Compare trends across categories
Visualize continuous relationships

Styling & Palettes

Customize visual appeal with themes and color palettes:

# Set 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")        # Minimal style with ticks

# Set color palette
sns.set_palette("pastel")     # Soft, muted colors
sns.set_palette("deep")       # Deep, rich colors
sns.set_palette("muted")       # Muted, balanced colors
sns.set_palette("bright")      # Bright, vibrant colors
sns.set_palette("dark")        # Dark colors
sns.set_palette("colorblind")  # Colorblind-friendly

# Custom palette
custom_palette = ["#FF6B6B", "#4ECDC4", "#45B7D1", "#FFA07A"]
sns.set_palette(custom_palette)

# Reset to default
sns.reset_orig()  # Reset to matplotlib defaults

Seaborn Categorical Plots

Categorical plots are used when one variable is categorical (like gender, day, or class). They help compare counts, averages, and spread across different groups.

barplot()

Displays the average of a numeric variable for each category:

# Basic bar plot
sns.barplot(data=tips, x='day', y='total_bill')
plt.title('Average Bill by Day')
plt.show()

# With hue (grouping)
sns.barplot(data=tips, x='day', y='total_bill', hue='sex')
plt.show()

# Horizontal bar plot
sns.barplot(data=tips, x='total_bill', y='day', orient='h')
plt.show()

# Custom estimator (e.g., median instead of mean)
sns.barplot(data=tips, x='day', y='total_bill', estimator=np.median)
plt.show()

countplot()

Shows the frequency of observations per category:

# Basic count plot
sns.countplot(data=tips, x='day')
plt.title('Frequency by Day')
plt.show()

# With hue
sns.countplot(data=tips, x='day', hue='sex')
plt.show()

# Horizontal
sns.countplot(data=tips, y='day')
plt.show()

# Order categories
sns.countplot(data=tips, x='day', order=['Thur', 'Fri', 'Sat', 'Sun'])
plt.show()

boxplot()

Visualizes median, quartiles, and outliers:

# Basic box plot
sns.boxplot(data=tips, x='day', y='total_bill')
plt.title('Bill Distribution by Day')
plt.show()

# With hue
sns.boxplot(data=tips, x='day', y='total_bill', hue='sex')
plt.show()

# Horizontal
sns.boxplot(data=tips, x='total_bill', y='day', orient='h')
plt.show()

# Customize
sns.boxplot(data=tips, x='day', y='total_bill', 
            palette='Set2', linewidth=2)
plt.show()

violinplot()

Combines boxplot with distribution shape for deeper insight:

# Basic violin plot
sns.violinplot(data=tips, x='day', y='total_bill')
plt.title('Distribution Shape by Day')
plt.show()

# With hue
sns.violinplot(data=tips, x='day', y='total_bill', hue='sex')
plt.show()

# Split violins
sns.violinplot(data=tips, x='day', y='total_bill', 
               hue='sex', split=True)
plt.show()

# Inner plot type
sns.violinplot(data=tips, x='day', y='total_bill', 
               inner='box')  # or 'quart', 'point', 'stick'
plt.show()

stripplot()

Displays all data points, even if they overlap:

# Basic strip plot
sns.stripplot(data=tips, x='day', y='total_bill')
plt.title('All Data Points by Day')
plt.show()

# With jitter (reduce overlap)
sns.stripplot(data=tips, x='day', y='total_bill', jitter=True)
plt.show()

# With hue
sns.stripplot(data=tips, x='day', y='total_bill', hue='sex')
plt.show()

# Customize
sns.stripplot(data=tips, x='day', y='total_bill', 
              size=4, alpha=0.5, palette='Set2')
plt.show()

swarmplot()

Similar to stripplot but prevents overlap of points:

# Basic swarm plot
sns.swarmplot(data=tips, x='day', y='total_bill')
plt.title('Non-Overlapping Points by Day')
plt.show()

# With hue
sns.swarmplot(data=tips, x='day', y='total_bill', hue='sex')
plt.show()

# Combined with violin plot
sns.violinplot(data=tips, x='day', y='total_bill', inner=None)
sns.swarmplot(data=tips, x='day', y='total_bill', 
              color='white', edgecolor='gray')
plt.show()

Practical Use Cases:

Compare customer spending by weekday
Analyze score distributions by gender
Visualize category frequencies in survey or sales data

Seaborn Distribution Plots

Distribution plots visualize how data values are spread, concentrated, or distributed across a range.

histplot()

Modern histogram with options for bins and KDE overlay:

# Basic histogram
sns.histplot(data=tips, x='total_bill', bins=30)
plt.title('Distribution of Total Bill')
plt.show()

# With KDE overlay
sns.histplot(data=tips, x='total_bill', bins=30, kde=True)
plt.show()

# With hue (compare multiple groups)
sns.histplot(data=tips, x='total_bill', hue='sex', bins=30, kde=True)
plt.show()

# Stacked histogram
sns.histplot(data=tips, x='total_bill', hue='sex', 
             bins=30, multiple='stack')
plt.show()

# Step histogram
sns.histplot(data=tips, x='total_bill', bins=30, 
             element='step', fill=False)
plt.show()

kdeplot()

Displays a smooth density curve representing data distribution:

# Basic KDE plot
sns.kdeplot(data=tips, x='total_bill')
plt.title('Density Curve of Total Bill')
plt.show()

# With shade
sns.kdeplot(data=tips, x='total_bill', shade=True)
plt.show()

# With hue
sns.kdeplot(data=tips, x='total_bill', hue='sex', shade=True)
plt.show()

# 2D KDE plot
sns.kdeplot(data=tips, x='total_bill', y='tip', shade=True)
plt.show()

# Cumulative distribution
sns.kdeplot(data=tips, x='total_bill', cumulative=True)
plt.show()

rugplot()

Adds small tick marks for individual observations:

# Basic rug plot
sns.rugplot(data=tips, x='total_bill')
plt.show()

# Combined with histogram and KDE
sns.histplot(data=tips, x='total_bill', bins=30, kde=True)
sns.rugplot(data=tips, x='total_bill', height=0.05, alpha=0.5)
plt.show()

Note on distplot()

distplot() is deprecated - replaced by histplot() and kdeplot():

# OLD (deprecated):
# sns.distplot(tips['total_bill'])

# NEW (use instead):
sns.histplot(data=tips, x='total_bill', kde=True)  # Histogram with KDE
# or
sns.kdeplot(data=tips, x='total_bill')  # Just KDE

Advanced Use with hue:

# Compare multiple groups
sns.histplot(data=tips, x='total_bill', hue='sex', 
             bins=30, kde=True, alpha=0.6)
plt.title('Distribution Comparison: Male vs Female Spending')
plt.show()

Practical Use Cases:

Analyze distribution of prices, customer tips, or sales amounts
Compare value distributions across categories for insights
Identify data patterns and outliers

Seaborn Regression & Mixed Plots

Regression and mixed plots visualize relationships between variables along with trend lines or combined plot types.

regplot()

Displays a scatterplot with a fitted regression line:

# Basic regression plot
sns.regplot(data=tips, x='total_bill', y='tip')
plt.title('Total Bill vs Tip with Regression Line')
plt.show()

# With confidence interval
sns.regplot(data=tips, x='total_bill', y='tip', ci=95)
plt.show()

# Without confidence interval
sns.regplot(data=tips, x='total_bill', y='tip', ci=None)
plt.show()

# Customize markers
sns.regplot(data=tips, x='total_bill', y='tip', 
            marker='+', scatter_kws={'s': 50, 'alpha': 0.5})
plt.show()

# Polynomial regression
sns.regplot(data=tips, x='total_bill', y='tip', order=2)
plt.show()

lmplot()

Similar to regplot but supports grouping (hue) and faceting:

# Basic lmplot
sns.lmplot(data=tips, x='total_bill', y='tip')
plt.show()

# With hue (grouping)
sns.lmplot(data=tips, x='total_bill', y='tip', hue='sex')
plt.show()

# With faceting (separate plots)
sns.lmplot(data=tips, x='total_bill', y='tip', 
           col='day', col_wrap=2)
plt.show()

# Multiple facets
sns.lmplot(data=tips, x='total_bill', y='tip', 
           row='sex', col='day')
plt.show()

# Customize
sns.lmplot(data=tips, x='total_bill', y='tip', 
           hue='sex', markers=['o', 's'], palette='Set1')
plt.show()

jointplot()

Combines scatter and histogram/KDE for joint and marginal distributions:

# Basic joint plot (scatter + histograms)
sns.jointplot(data=tips, x='total_bill', y='tip')
plt.show()

# With KDE
sns.jointplot(data=tips, x='total_bill', y='tip', kind='kde')
plt.show()

# With regression
sns.jointplot(data=tips, x='total_bill', y='tip', kind='reg')
plt.show()

# With hex bins
sns.jointplot(data=tips, x='total_bill', y='tip', kind='hex')
plt.show()

# With scatter and KDE
sns.jointplot(data=tips, x='total_bill', y='tip', 
              kind='scatter', marginal_kws=dict(bins=30, kde=True))
plt.show()

pairplot()

Shows pairwise relationships among multiple variables in a dataset:

# Basic pair plot
iris = sns.load_dataset('iris')
sns.pairplot(iris)
plt.show()

# With hue (grouping)
sns.pairplot(iris, hue='species')
plt.show()

# Customize diagonal
sns.pairplot(iris, hue='species', diag_kind='kde')
plt.show()

# Select specific variables
sns.pairplot(iris, vars=['sepal_length', 'sepal_width', 'petal_length'], 
             hue='species')
plt.show()

# Customize markers
sns.pairplot(iris, hue='species', markers=['o', 's', 'D'])
plt.show()

Practical Use Cases:

Analyze correlations (e.g., total bill vs tip)
Visualize gender-based trends
Explore relationships among multiple numeric features in datasets like Iris

Seaborn Matrix & Styling Plots

Matrix or grid plots are great for exploring correlations, patterns, and relationships within datasets.

heatmap()

Displays correlations or pivot tables as colored grids:

# Correlation heatmap
correlation = tips.select_dtypes(include=[np.number]).corr()
plt.figure(figsize=(10, 8))
sns.heatmap(correlation, annot=True, cmap='coolwarm', center=0)
plt.title('Correlation Heatmap')
plt.show()

# Pivot table heatmap
pivot = tips.pivot_table(values='total_bill', index='day', columns='time')
sns.heatmap(pivot, annot=True, fmt='.2f', cmap='YlOrRd')
plt.title('Total Bill Heatmap')
plt.show()

# Customize
sns.heatmap(correlation, annot=True, fmt='.2f', 
            cmap='coolwarm', center=0,
            square=True, linewidths=1, cbar_kws={'label': 'Correlation'})
plt.show()

# Without annotations
sns.heatmap(correlation, annot=False, cmap='viridis')
plt.show()

clustermap()

Adds clustering to heatmaps for grouping similar variables or patterns:

# Basic clustermap
sns.clustermap(correlation, annot=True, cmap='coolwarm', center=0)
plt.show()

# With different metric
sns.clustermap(correlation, metric='euclidean', method='ward')
plt.show()

# Customize
sns.clustermap(correlation, annot=True, fmt='.2f',
               cmap='coolwarm', figsize=(10, 8),
               row_cluster=True, col_cluster=True)
plt.show()

Styling Options

Control global styling for all plots:

# Set style
sns.set_style("white")        # White background
sns.set_style("dark")         # Dark background
sns.set_style("whitegrid")    # White with grid (default)
sns.set_style("darkgrid")     # Dark with grid
sns.set_style("ticks")        # Minimal with ticks

# Set palette
sns.set_palette("Set2")       # Qualitative palette
sns.set_palette("pastel")     # Soft colors
sns.set_palette("deep")       # Deep colors
sns.set_palette("muted")      # Muted colors
sns.set_palette("bright")     # Bright colors
sns.set_palette("dark")       # Dark colors
sns.set_palette("colorblind") # Colorblind-friendly

# Set context (affects scale)
sns.set_context("paper")      # Smallest (for papers)
sns.set_context("notebook")   # Default
sns.set_context("talk")       # Larger (for presentations)
sns.set_context("poster")     # Largest (for posters)

# Combine settings
sns.set_style("whitegrid")
sns.set_palette("Set2")
sns.set_context("notebook")

# Reset to defaults
sns.reset_orig()

Use Cases:

Analyze correlations between variables
Highlight variable groupings
Create visually consistent plots with global themes

Seaborn Statistical Visualizations

Distribution Plots

# Load sample data
tips = sns.load_dataset('tips')

# Distribution plot
plt.figure(figsize=(10, 6))
sns.histplot(data=tips, x='total_bill', kde=True, bins=30)
plt.title('Distribution of Total Bill')
plt.show()

# KDE plot
plt.figure(figsize=(10, 6))
sns.kdeplot(data=tips, x='total_bill', shade=True)
plt.title('KDE Plot')
plt.show()

Categorical Plots

# Bar plot
plt.figure(figsize=(10, 6))
sns.barplot(data=tips, x='day', y='total_bill', hue='sex')
plt.title('Average Bill by Day and Gender')
plt.show()

# Box plot
plt.figure(figsize=(10, 6))
sns.boxplot(data=tips, x='day', y='total_bill', hue='sex')
plt.title('Bill Distribution by Day')
plt.show()

# Violin plot
plt.figure(figsize=(10, 6))
sns.violinplot(data=tips, x='day', y='total_bill')
plt.title('Violin Plot')
plt.show()

Relationship Plots

# Scatter plot
plt.figure(figsize=(10, 6))
sns.scatterplot(data=tips, x='total_bill', y='tip', hue='sex', size='size')
plt.title('Tip vs Total Bill')
plt.show()

# Regression plot
plt.figure(figsize=(10, 6))
sns.regplot(data=tips, x='total_bill', y='tip')
plt.title('Regression Plot')
plt.show()

# Pair plot (multiple relationships)
sns.pairplot(tips, hue='sex', diag_kind='kde')
plt.show()

Heatmap

# Correlation heatmap
correlation = tips.select_dtypes(include=[np.number]).corr()
plt.figure(figsize=(10, 8))
sns.heatmap(correlation, annot=True, cmap='coolwarm', center=0)
plt.title('Correlation Heatmap')
plt.show()

# Pivot table heatmap
pivot = tips.pivot_table(values='total_bill', index='day', columns='time')
plt.figure(figsize=(10, 6))
sns.heatmap(pivot, annot=True, fmt='.2f', cmap='YlOrRd')
plt.title('Heatmap of Total Bill')
plt.show()

# Create facet grid
g = sns.FacetGrid(tips, col='day', row='sex', height=4)
g.map(sns.scatterplot, 'total_bill', 'tip')
g.add_legend()
plt.show()

Saving Figures

Saving Plots

# Create plot
plt.figure(figsize=(10, 6))
plt.plot(np.linspace(0, 10, 100), np.sin(np.linspace(0, 10, 100)))
plt.title('Sine Wave')

# Save with different formats
plt.savefig('plot.png', dpi=300, bbox_inches='tight')  # PNG
plt.savefig('plot.pdf', bbox_inches='tight')           # PDF
plt.savefig('plot.svg', bbox_inches='tight')          # SVG
plt.savefig('plot.jpg', dpi=300, quality=95)          # JPEG

# With transparent background
plt.savefig('plot.png', transparent=True, dpi=300)

plt.show()

High-Quality Figures

# For publications
fig, ax = plt.subplots(figsize=(8, 6), dpi=300)
ax.plot(x, y, linewidth=2)
ax.set_xlabel('X', fontsize=14)
ax.set_ylabel('Y', fontsize=14)
ax.set_title('Publication Quality Figure', fontsize=16)
plt.tight_layout()
plt.savefig('publication_figure.png', dpi=300, bbox_inches='tight')
plt.show()

Practice Exercises

Exercise 1: Basic Line Plot

Task: Create a line plot of y = x² from -5 to 5 with proper labels and grid.

Solution:

x = np.linspace(-5, 5, 100)
y = x**2

plt.figure(figsize=(10, 6))
plt.plot(x, y, linewidth=2, color='blue')
plt.xlabel('X', fontsize=12)
plt.ylabel('Y = X²', fontsize=12)
plt.title('Quadratic Function', fontsize=14)
plt.grid(True, alpha=0.3)
plt.show()

Exercise 2: Multiple Subplots

Task: Create a 2x2 subplot with different plot types.

Solution:

fig, axes = plt.subplots(2, 2, figsize=(12, 10))

# Line plot
x = np.linspace(0, 10, 100)
axes[0, 0].plot(x, np.sin(x))
axes[0, 0].set_title('Sine Wave')

# Scatter plot
axes[0, 1].scatter(np.random.randn(100), np.random.randn(100))
axes[0, 1].set_title('Scatter Plot')

# Bar plot
axes[1, 0].bar(['A', 'B', 'C'], [10, 20, 30])
axes[1, 0].set_title('Bar Plot')

# Histogram
axes[1, 1].hist(np.random.normal(0, 1, 1000), bins=30)
axes[1, 1].set_title('Histogram')

plt.tight_layout()
plt.show()

Exercise 3: Data Visualization

Task: Load a dataset and create visualizations showing distributions and relationships.

Solution:

# Using tips dataset
tips = sns.load_dataset('tips')

# Create figure with subplots
fig, axes = plt.subplots(2, 2, figsize=(14, 10))

# Distribution
sns.histplot(data=tips, x='total_bill', kde=True, ax=axes[0, 0])
axes[0, 0].set_title('Distribution of Total Bill')

# Box plot
sns.boxplot(data=tips, x='day', y='total_bill', ax=axes[0, 1])
axes[0, 1].set_title('Bill by Day')

# Scatter plot
sns.scatterplot(data=tips, x='total_bill', y='tip', hue='sex', ax=axes[1, 0])
axes[1, 0].set_title('Tip vs Bill')

# Bar plot
sns.barplot(data=tips, x='day', y='total_bill', ax=axes[1, 1])
axes[1, 1].set_title('Average Bill by Day')

plt.tight_layout()
plt.show()

Exercise 4: Correlation Heatmap

Task: Create a correlation heatmap for numeric columns in a dataset.

Solution:

# Calculate correlation
df = pd.DataFrame({
    'A': np.random.randn(100),
    'B': np.random.randn(100),
    'C': np.random.randn(100),
    'D': np.random.randn(100)
})

correlation = df.corr()

plt.figure(figsize=(10, 8))
sns.heatmap(correlation, annot=True, cmap='coolwarm', center=0, 
            square=True, linewidths=1)
plt.title('Correlation Heatmap')
plt.show()

Key Takeaways

Matplotlib: Full control, use for custom plots
Seaborn: Beautiful defaults, use for statistical plots
Subplots: Organize multiple plots
Customization: Always label axes and add titles
Save high-quality: Use high DPI for publications

Common Patterns

Pattern 1: EDA Visualization

def plot_eda(df):
    """Create comprehensive EDA visualizations"""
    fig, axes = plt.subplots(2, 2, figsize=(14, 10))
    
    # Distribution
    axes[0, 0].hist(df['column'], bins=30)
    axes[0, 0].set_title('Distribution')
    
    # Box plot
    axes[0, 1].boxplot(df['column'])
    axes[0, 1].set_title('Box Plot')
    
    # Scatter
    axes[1, 0].scatter(df['x'], df['y'])
    axes[1, 0].set_title('Scatter Plot')
    
    # Correlation
    sns.heatmap(df.corr(), annot=True, ax=axes[1, 1])
    axes[1, 1].set_title('Correlation')
    
    plt.tight_layout()
    plt.show()

Plotly and Dash for Interactive Visualizations

Plotly is a powerful library for creating interactive, publication-quality visualizations. Dash is a framework built on Plotly for building analytical web applications.

Why Plotly?

Advantages:

Interactive: Hover, zoom, pan, and filter
Publication Quality: Professional-looking plots
Web-Ready: Works in Jupyter notebooks and web browsers
Multiple Backends: Python, R, JavaScript
Rich Features: 3D plots, animations, subplots

Installation

pip install plotly dash

Plotly Express (Simple API)

Plotly Express provides a simple, high-level interface for creating plots.

Basic Plots

import plotly.express as px
import pandas as pd
import numpy as np

# Create sample data
df = pd.DataFrame({
    'x': np.random.randn(100),
    'y': np.random.randn(100),
    'category': np.random.choice(['A', 'B', 'C'], 100),
    'size': np.random.rand(100) * 100
})

# Scatter plot
fig = px.scatter(df, x='x', y='y', color='category', size='size',
                 title='Interactive Scatter Plot',
                 labels={'x': 'X Axis', 'y': 'Y Axis'})
fig.show()

# Line plot
df_time = pd.DataFrame({
    'date': pd.date_range('2024-01-01', periods=100),
    'value': np.cumsum(np.random.randn(100))
})
fig = px.line(df_time, x='date', y='value', title='Time Series')
fig.show()

# Bar chart
df_bar = pd.DataFrame({
    'category': ['A', 'B', 'C', 'D'],
    'value': [10, 20, 15, 25]
})
fig = px.bar(df_bar, x='category', y='value', title='Bar Chart')
fig.show()

# Histogram
fig = px.histogram(df, x='x', nbins=30, title='Histogram')
fig.show()

# Box plot
fig = px.box(df, x='category', y='y', title='Box Plot')
fig.show()

Advanced Plotly Express

# 3D Scatter
fig = px.scatter_3d(df, x='x', y='y', z='size', color='category')
fig.show()

# Faceted plots
fig = px.scatter(df, x='x', y='y', facet_col='category', 
                 title='Faceted Scatter Plot')
fig.show()

# Heatmap
correlati>'x', 'y', 'size']].corr()
fig = px.imshow(correlation_matrix, text_auto=True, aspect="auto",
                title='Correlation Heatmap')
fig.show()

Plotly Graph Objects (Advanced Control)

For more control, use Plotly Graph Objects:

import plotly.graph_objects as go
from plotly.subplots import make_subplots

# Create figure
fig = go.Figure()

# Add traces
fig.add_trace(go.Scatter(
    x=[1, 2, 3, 4],
    y=[10, 11, 12, 13],
    mode='lines+markers',
    name='Line 1'
))

# Update layout
fig.update_layout(
    title='Custom Plot',
    xaxis_title='X Axis',
    yaxis_title='Y Axis',
    hovermode='x unified'
)

fig.show()

Dash for Web Applications

Dash allows you to build interactive web applications with Plotly.

Basic Dash App

import dash
from dash import dcc, html
import plotly.express as px
import pandas as pd

# Create Dash app
app = dash.Dash(__name__)

# Create figure
fig = px.scatter(df, x='x', y='y', color='category')

# Define app layout
app.layout = html.Div([
    html.H1("My Dash App"),
    dcc.Graph(figure=fig)
])

# Run app
if __name__ == '__main__':
    app.run_server(debug=True)

Key Takeaways for Plotly and Dash

Plotly Express: Simple API for quick interactive plots
Plotly Graph Objects: Advanced control and customization
Dash: Build interactive web applications
Interactivity: Hover, zoom, pan, filter built-in
Deployment: Deploy Dash apps to cloud platforms

Next Steps

Practice creating different plot types
Work with real datasets
Experiment with customization
Try Plotly for interactive visualizations
Build a Dash dashboard for your project
Move to 02-introduction-to-ml for ML concepts

Try next: Redraw one chart so a stranger can answer the business question in five seconds.

Recall ::

What should a good plot answer before you decorate it?

One clear question about the data a stranger can read in a few seconds.

Data Visualization

ML toolbox curriculum map (this guide)

Table of Contents

Introduction

What is Matplotlib?

What is Seaborn?

Installation

Matplotlib Basics

Simple Plot

Multiple Lines

Figure and Axes

Figure and Axes Structure

Components

Understanding the Hierarchy

Common Plot Types

Line Plot

Scatter Plot

Bar Plot

Histogram

Box Plot

Pie Chart

Customizing Plots

Colors and Styles

Axis Customization

Text and Annotations

Subplots and Multiple Plots

Subplots

GridSpec (Advanced Layout)

Seaborn Introduction

What is Seaborn?

Installation and Setup

Seaborn Relational Plots

scatterplot()

lineplot()

Styling & Palettes

Seaborn Categorical Plots

barplot()

countplot()

boxplot()

violinplot()

stripplot()

swarmplot()

Seaborn Distribution Plots

histplot()

kdeplot()

rugplot()

Note on distplot()

Seaborn Regression & Mixed Plots

regplot()

lmplot()

jointplot()

pairplot()

Seaborn Matrix & Styling Plots

heatmap()

clustermap()

Styling Options

Seaborn Statistical Visualizations

Distribution Plots

Categorical Plots

Relationship Plots

Heatmap

Facet Grid

Saving Figures

Saving Plots

High-Quality Figures

Practice Exercises

Exercise 1: Basic Line Plot

Exercise 2: Multiple Subplots

Exercise 3: Data Visualization

Exercise 4: Correlation Heatmap

Key Takeaways

Common Patterns

Pattern 1: EDA Visualization

Plotly and Dash for Interactive Visualizations

Why Plotly?

Installation

Plotly Express (Simple API)

Basic Plots

Advanced Plotly Express

Plotly Graph Objects (Advanced Control)