Run this notebook: Open in Colab Open in Kaggle

Data Visualization: From Exploratory Plots to Publication-Quality Figures¶

Matplotlib fundamentals, seaborn statistical plots, and plotly interactive charts — with a focus on making visualizations that actually communicate insights.

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

# Style
plt.style.use('seaborn-v0_8-whitegrid')
sns.set_palette('husl')

# Datasets
tips = sns.load_dataset('tips')
flights = sns.load_dataset('flights')
iris = sns.load_dataset('iris')

print('Datasets loaded: tips, flights, iris')

1. Matplotlib Subplots — Full Control¶

fig, axes = plt.subplots(2, 2, figsize=(14, 10))
fig.suptitle('Tips Dataset: Four Perspectives', fontsize=16, fontweight='bold')

# 1. Total bill distribution
axes[0, 0].hist(tips['total_bill'], bins=25, color='steelblue', edgecolor='white', alpha=0.8)
axes[0, 0].axvline(tips['total_bill'].mean(), color='red', linestyle='--', label=f'Mean: ${tips["total_bill"].mean():.2f}')
axes[0, 0].axvline(tips['total_bill'].median(), color='orange', linestyle='--', label=f'Median: ${tips["total_bill"].median():.2f}')
axes[0, 0].set_title('Total Bill Distribution')
axes[0, 0].set_xlabel('Total Bill ($)')
axes[0, 0].legend()

# 2. Tip % by day
tips['tip_pct'] = tips['tip'] / tips['total_bill'] * 100
day_order = ['Thur', 'Fri', 'Sat', 'Sun']
bp = axes[0, 1].boxplot(
    [tips[tips['day'] == d]['tip_pct'].values for d in day_order],
    labels=day_order,
    patch_artist=True,
    medianprops={'color': 'red', 'linewidth': 2}
)
colors = ['#3498db', '#e74c3c', '#2ecc71', '#f39c12']
for patch, color in zip(bp['boxes'], colors):
    patch.set_facecolor(color)
    patch.set_alpha(0.7)
axes[0, 1].set_title('Tip % Distribution by Day')
axes[0, 1].set_ylabel('Tip %')

# 3. Scatter: bill vs tip, colored by sex
for sex, color, marker in [('Male', 'steelblue', 'o'), ('Female', 'salmon', 's')]:
    subset = tips[tips['sex'] == sex]
    axes[1, 0].scatter(subset['total_bill'], subset['tip'],
                       c=color, marker=marker, alpha=0.6, s=50, label=sex)
m, b = np.polyfit(tips['total_bill'], tips['tip'], 1)
x_line = np.linspace(tips['total_bill'].min(), tips['total_bill'].max(), 100)
axes[1, 0].plot(x_line, m * x_line + b, 'k--', alpha=0.5, label=f'Trend (r²={np.corrcoef(tips["total_bill"], tips["tip"])[0,1]**2:.2f})')
axes[1, 0].set_title('Total Bill vs Tip')
axes[1, 0].set_xlabel('Total Bill ($)')
axes[1, 0].set_ylabel('Tip ($)')
axes[1, 0].legend()

# 4. Stacked bar: smoker × time
counts = tips.groupby(['time', 'smoker']).size().unstack(fill_value=0)
counts.plot(kind='bar', stacked=True, ax=axes[1, 1],
            color=['#e74c3c', '#3498db'], alpha=0.8, edgecolor='white')
axes[1, 1].set_title('Diners by Meal Time & Smoking')
axes[1, 1].set_xlabel('Time')
axes[1, 1].tick_params(axis='x', rotation=0)

plt.tight_layout()
plt.show()

2. Seaborn — Statistical Visualizations¶

fig, axes = plt.subplots(1, 3, figsize=(18, 5))

# Violin plot — richer than boxplot
sns.violinplot(data=tips, x='day', y='tip_pct', hue='sex',
               split=True, inner='quart', ax=axes[0],
               order=['Thur', 'Fri', 'Sat', 'Sun'])
axes[0].set_title('Tip % by Day & Sex (Violin)')
axes[0].set_ylabel('Tip %')

# Heatmap — flights data (months × year)
flights_pivot = flights.pivot(index='month', columns='year', values='passengers')
sns.heatmap(flights_pivot, annot=True, fmt='d', cmap='YlOrRd',
            ax=axes[1], linewidths=0.5, cbar_kws={'label': 'Passengers'})
axes[1].set_title('Monthly Airline Passengers (1949-1960)')

# Pair plot subset (use FacetGrid for custom pair plots inline)
sns.scatterplot(data=iris, x='sepal_length', y='petal_length',
                hue='species', style='species', s=80, ax=axes[2])
axes[2].set_title('Iris: Sepal vs Petal Length')

plt.tight_layout()
plt.show()

# FacetGrid — small multiples
g = sns.FacetGrid(tips, col='time', row='sex', height=4, aspect=1.2)
g.map_dataframe(sns.histplot, x='total_bill', bins=20, kde=True)
g.set_axis_labels('Total Bill ($)', 'Count')
g.set_titles(col_template='{col_name}', row_template='{row_name}')
g.add_legend()
plt.suptitle('Bill Distribution by Meal × Sex (Small Multiples)', y=1.02, fontsize=13)
plt.tight_layout()
plt.show()

3. Publication-Quality Figures¶

# Techniques for making figures presentable

COLORS = {'primary': '#2c3e50', 'accent': '#e74c3c', 'highlight': '#f39c12', 'muted': '#95a5a6'}

fig, ax = plt.subplots(figsize=(10, 6))

# Monthly trend
monthly = flights.groupby('year')['passengers'].sum().reset_index()
ax.fill_between(monthly['year'], monthly['passengers'],
                alpha=0.15, color=COLORS['primary'])
ax.plot(monthly['year'], monthly['passengers'],
        color=COLORS['primary'], linewidth=2.5, marker='o', markersize=6)

# Highlight peak year
peak = monthly.loc[monthly['passengers'].idxmax()]
ax.scatter(peak['year'], peak['passengers'],
           color=COLORS['accent'], s=150, zorder=5)
ax.annotate(f'Peak: {peak["passengers"]:,}',
            xy=(peak['year'], peak['passengers']),
            xytext=(peak['year'] - 2, peak['passengers'] - 5000),
            fontsize=11, color=COLORS['accent'],
            arrowprops={'arrowstyle': '->', 'color': COLORS['accent']})

# Polish
ax.set_title('Annual Airline Passengers: 1949–1960', fontsize=15, fontweight='bold', pad=15)
ax.set_xlabel('Year', fontsize=12)
ax.set_ylabel('Total Passengers', fontsize=12)
ax.yaxis.set_major_formatter(plt.FuncFormatter(lambda x, _: f'{x/1000:.0f}K'))
ax.spines[['top', 'right']].set_visible(False)  # Remove top/right borders
ax.grid(axis='y', alpha=0.4)

plt.tight_layout()
plt.show()

4. Plotly — Interactive Charts¶

# !pip install plotly
try:
    import plotly.express as px
    import plotly.graph_objects as go
    from plotly.subplots import make_subplots
    
    # Interactive scatter
    fig = px.scatter(
        tips, x='total_bill', y='tip',
        color='sex', size='tip_pct', symbol='time',
        hover_data=['day', 'size'],
        title='Tips: Interactive Explorer (hover for details)',
        labels={'total_bill': 'Total Bill ($)', 'tip': 'Tip ($)'}
    )
    fig.show()
    
    # Interactive heatmap
    fig2 = px.imshow(
        flights_pivot,
        title='Airline Passengers Heatmap (interactive)',
        color_continuous_scale='YlOrRd',
        aspect='auto'
    )
    fig2.show()

except ImportError:
    print('Install plotly: pip install plotly')
    print('Example code shown above — run after installing.')

Visualization Cheat Sheet¶

Data type / Goal                  → Best chart
──────────────────────────────────────────────
Single continuous variable        → histogram + KDE
Distribution comparison           → violin plot or overlaid KDE
Category comparison               → bar chart with error bars
Trend over time                   → line chart with shaded area
Two continuous variables          → scatter with regression line
Many variables pairwise           → pairplot (seaborn)
Matrix of values                  → heatmap
Part of whole                     → stacked bar (not pie!)
Multiple group distributions      → FacetGrid small multiples
Interactive exploration           → plotly express

Exercises¶

Recreate the matplotlib multi-panel figure using seaborn API only.
Load sns.load_dataset('penguins') and create a pairplot colored by species.
Build a Plotly animated scatter plot showing flight passengers growing over time using animation_frame='year'.
Create a dashboard-style 3×2 figure for the tips dataset with a consistent color scheme.
Use matplotlib.gridspec.GridSpec to create an asymmetric layout (large scatter + two small histograms as marginals).