Matplotlib in Python

Introduction to Matplotlib

Matplotlib is Python's foundational plotting library — the engine under almost every visualisation in the scientific and data science ecosystem. Seaborn, pandas .plot(), and dozens of other libraries are all built on top of it. Understanding matplotlib directly gives you complete control over every pixel of every chart you produce.

The Anatomy of a Matplotlib Figure

Setup and Imports

import matplotlib.pyplot as plt
import matplotlib.ticker as mticker
import numpy as np
import pandas as pd

# Apply a style sheet globally
plt.style.use('seaborn-v0_8-darkgrid')   # clean dark grid look

# Set default figure size and resolution
plt.rcParams['figure.figsize']    = (10, 5)
plt.rcParams['figure.dpi']        = 120
plt.rcParams['axes.titlesize']    = 14
plt.rcParams['axes.labelsize']    = 12
plt.rcParams['xtick.labelsize']   = 10
plt.rcParams['ytick.labelsize']   = 10
plt.rcParams['legend.fontsize']   = 10
plt.rcParams['lines.linewidth']   = 2

Line Chart — Trends Over Time

The line chart is matplotlib's most used plot. It connects data points in sequence — ideal for time series, continuous functions, and any data where order matters. Every visual property of the line (colour, width, style, markers) is independently controllable.

fig, ax = plt.subplots(figsize=(10, 5))

# Multiple lines with different styles
ax.plot(months, revenue,
       color='#60a5fa', linewidth=2, marker='o', markersize=5,
       label='Revenue')

ax.plot(months, target,
       color='#f59e0b', linewidth=1.5, linestyle='--',
       label='Target')

# Fill the area between lines
ax.fill_between(months, revenue, target,
               where=[r > t for r, t in zip(revenue, target)],
               alpha=0.15, color='#34d399', label='Above target')

ax.set_title('Monthly Revenue vs Target', fontsize=14, fontweight='bold', pad=12)
ax.set_xlabel('Month')
ax.set_ylabel('Revenue (₹ thousands)')
ax.legend(framealpha=0.3)
ax.yaxis.set_major_formatter(mticker.FuncFormatter(lambda x, _: f'₹{x:.0f}k'))
plt.tight_layout()
plt.show()

Bar Chart — Comparing Categories

Bar charts compare a numeric measure across discrete categories. Matplotlib supports vertical bars, horizontal bars, grouped bars, and stacked bars — each with independent colour and width control per bar.

fig, axes = plt.subplots(1, 2, figsize=(12, 5))

# ── Grouped bar chart ──────────────────────────────
x = np.arange(len(categories))
width = 0.35

bars1 = axes[0].bar(x - width/2, male_avg,   width, label='Male',   color='#60a5fa', alpha=0.85)
bars2 = axes[0].bar(x + width/2, female_avg, width, label='Female', color='#f87171', alpha=0.85)

# Add value labels on top of each bar
axes[0].bar_label(bars1, fmt='₹%.0fk', padding=3, fontsize=9)
axes[0].bar_label(bars2, fmt='₹%.0fk', padding=3, fontsize=9)

axes[0].set_xticks(x)
axes[0].set_xticklabels(categories, rotation=30, ha='right')
axes[0].set_title('Avg Spend by Category & Gender')
axes[0].legend()

# ── Horizontal bar chart (sorted) ──────────────────
sorted_idx = np.argsort(total_revenue)
axes[1].barh(np.array(categories)[sorted_idx],
             np.array(total_revenue)[sorted_idx],
             color='#f59e0b', alpha=0.85)
axes[1].set_title('Total Revenue by Category')
plt.tight_layout()
plt.show()

Scatter Plot — Relationships Between Variables

Scatter plots reveal correlations, clusters, and outliers. In matplotlib, the scatter() function encodes up to four dimensions simultaneously: x position, y position, colour (third variable), and size (fourth variable).

fig, ax = plt.subplots(figsize=(9, 6))

scatter = ax.scatter(
    df['age'], df['purchase_amount'],
    c=df['rating'],          # colour encodes a 3rd variable
    s=df['delivery_days'] * 8, # size encodes a 4th variable
    cmap='viridis',
    alpha=0.6, edgecolors='white', linewidths=0.4
)

# Colourbar for the 3rd dimension
cbar = plt.colorbar(scatter, ax=ax)
cbar.set_label('Customer Rating', fontsize=11)

# Annotate a specific outlier point
ax.annotate('High-value outlier',
            xy=(58, 24800), xytext=(45, 23000),
            arrowprops=dict(arrowstyle='->', color='#f87171'),
            fontsize=9, color='#f87171')

ax.set_xlabel('Customer Age')
ax.set_ylabel('Purchase Amount (₹)')
ax.set_title('Age vs Purchase — colour=rating, size=delivery days')
plt.tight_layout()
plt.show()

Histogram & KDE — Visualising Distributions

Matplotlib's ax.hist() is the foundation for distribution analysis. Combined with a manually computed KDE line, it gives you the full picture of a variable's shape — skewness, modality, and tail weight — all in one chart.

from scipy.stats import gaussian_kde

fig, axes = plt.subplots(1, 2, figsize=(12, 5))

# ── Left: histogram + KDE ──────────────────────────
n, bins, patches = axes[0].hist(
    data, bins=35, color='#60a5fa', alpha=0.6,
    edgecolor='white', linewidth=0.4, density=True
)

# Overlay KDE curve
kde = gaussian_kde(data)
x_range = np.linspace(data.min(), data.max(), 300)
axes[0].plot(x_range, kde(x_range), color='#f59e0b', linewidth=2, label='KDE')

# Colour bars by region (below mean = blue, above = amber)
mean_val = data.mean()
for patch, left_edge in zip(patches, bins):
    if left_edge > mean_val:
        patch.set_facecolor('#f59e0b')

axes[0].axvline(mean_val, color='#f59e0b', linestyle='--', label=f'Mean={mean_val:.0f}')
axes[0].axvline(np.median(data), color='#34d399', linestyle='--', label='Median')
axes[0].legend()
axes[0].set_title('Distribution of Purchase Amount')

# ── Right: overlapping histograms by group ──────────
for group, colour in zip(groups, ['#60a5fa', '#f87171', '#34d399']):
    axes[1].hist(group['data'], bins=25, alpha=0.5,
                color=colour, label=group['label'], density=True)
axes[1].legend()
axes[1].set_title('Distribution by Income Bracket')
plt.tight_layout()
plt.show()

Subplots — Multiple Charts in One Figure

The plt.subplots() function creates a grid of axes objects. This is the core layout tool in matplotlib — it lets you create dashboards, comparison panels, and multi-panel analysis figures with precise shared axis control.

# 2×2 grid of subplots with shared x-axis
fig, axes = plt.subplots(2, 2, figsize=(12, 8),
                         sharex=False, sharey=False)

# Access individual axes
ax_line  = axes[0, 0]   # top-left
ax_bar   = axes[0, 1]   # top-right
ax_hist  = axes[1, 0]   # bottom-left
ax_scat  = axes[1, 1]   # bottom-right

# Add a shared super-title for the whole figure
fig.suptitle('Sales Dashboard — Q4 2024', fontsize=16, fontweight='bold', y=1.01)

# Remove unused axes: axes[1,2].set_visible(False)

# Adjust spacing between subplots
plt.tight_layout(pad=2.0)
plt.show()

Heatmap with imshow() — Correlation & Pivot Tables

Matplotlib's ax.imshow() renders any 2D array as a coloured grid — the foundation for correlation heatmaps, pivot table visualisations, and confusion matrices. Unlike seaborn's heatmap(), the matplotlib version gives you full control over colormap, cell annotations, and axis formatting.

fig, ax = plt.subplots(figsize=(7, 6))

# Render the correlation matrix as an image
im = ax.imshow(corr_matrix, cmap='coolwarm', vmin=-1, vmax=1, aspect='auto')

# Add colourbar
plt.colorbar(im, ax=ax, fraction=0.046, pad=0.04)

# Annotate every cell with the correlation value
for i in range(n):
    for j in range(n):
        val = corr_matrix[i, j]
        text_color = 'white' if abs(val) > 0.6 else 'black'
        ax.text(j, i, f'{val:.2f}', ha='center', va='center',
               fontsize=11, color=text_color, fontweight='bold')

ax.set_xticks(range(n)); ax.set_xticklabels(feature_names, rotation=45, ha='right')
ax.set_yticks(range(n)); ax.set_yticklabels(feature_names)
ax.set_title('Correlation Matrix (Pearson)', pad=14)
plt.tight_layout()
plt.show()

Boxplot & Violin Plot — Distribution Shape

Matplotlib's ax.boxplot() and ax.violinplot() expose every visual detail of a distribution: median, quartiles, whiskers, and outliers. The violin plot extends this by showing the full distribution shape via a KDE on each side.

fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(12, 5))

# ── Styled boxplot ─────────────────────────────────
bp = ax1.boxplot(
    data_by_category,
    patch_artist=True,        # fills boxes with colour
    notch=True,              # notch = 95% CI around median
    vert=True,
    widths=0.5
)
# Style each box individually
colors = ['#60a5fa', '#34d399', '#f59e0b', '#a78bfa', '#f87171']
for patch, color in zip(bp['boxes'], colors):
    patch.set_facecolor(color)
    patch.set_alpha(0.6)

# ── Violin plot ─────────────────────────────────────
vp = ax2.violinplot(
    data_by_category,
    showmedians=True, showextrema=True
)
for body, color in zip(vp['bodies'], colors):
    body.set_facecolor(color)
    body.set_alpha(0.5)

plt.tight_layout()
plt.show()

Styling & Saving Publication-Quality Figures

Matplotlib's style system and fine-grained control over every element makes it the gold standard for publication figures. Journals, conference papers, and reports all demand specific font sizes, exact figure dimensions, and lossless export — matplotlib handles all of these.

# ── List all available styles ───────────────────────
print(plt.style.available)

# ── Apply a style ───────────────────────────────────
plt.style.use('seaborn-v0_8-whitegrid')  # clean white
plt.style.use('dark_background')         # pure dark
plt.style.use('ggplot')                   # R-style
plt.style.use('bmh')                      # Bayesian methods

# ── Custom rcParams for publication ─────────────────
plt.rcParams.update({
    'font.family':      'DejaVu Sans',
    'figure.dpi':       150,
    'savefig.dpi':      300,
    'axes.spines.top':  False,   # remove top spine
    'axes.spines.right':False,   # remove right spine
    'axes.grid':        True,
    'grid.alpha':       0.3,
})

# ── Saving figures ───────────────────────────────────
fig.savefig('chart.png',  dpi=300, bbox_inches='tight')
fig.savefig('chart.pdf',  bbox_inches='tight')         # vector PDF
fig.savefig('chart.svg',  bbox_inches='tight')         # scalable SVG
fig.savefig('chart.eps',  bbox_inches='tight')         # for LaTeX

Golden Rules of Matplotlib