PY-L4-33 · Matplotlib — Bar Charts

Learning Goals

3 min

Install matplotlib; import matplotlib.pyplot as plt.
Draw a bar chart with plt.bar(x, y).
Title, axis labels, rotated x-ticks.
Annotate bars with their values.
Save the figure to PNG.

Warm-Up · 4 Lines, 1 Chart

5 min

pip install matplotlib

import matplotlib.pyplot as plt

products = ["Pancake", "Cocoa", "Rice", "Tea"]
revenue  = [120, 95, 180, 60]

plt.bar(products, revenue)
plt.show()

You'll get a window with four blue bars. It works. Now we make it presentable.

Today's big idea

Default matplotlib is ugly on purpose — it's the canvas you customise. Five extra lines (title, labels, color, rotation, value text) turn the warm-up plot into something you'd ship.

New Concept · plt.bar & the Customisations

14 min

Pyplot vs OO

Two styles exist; both work. We'll use OO (object-oriented) — it scales to subplots:

import matplotlib.pyplot as plt

fig, ax = plt.subplots(figsize=(8, 5))
ax.bar(products, revenue)
plt.show()

Title + axis labels

ax.set_title("Revenue by product · May 2026")
ax.set_xlabel("Product")
ax.set_ylabel("Revenue ($)")

Colour and edge

bars = ax.bar(products, revenue,
              color="#4a90e2", edgecolor="black", linewidth=0.5)

Rotated x labels (when names are long)

ax.tick_params(axis="x", rotation=30)
# or
plt.xticks(rotation=30, ha="right")

Value annotations on each bar

for bar in bars:
    h = bar.get_height()
    ax.text(bar.get_x() + bar.get_width() / 2, h + 2,
            f"{h}", ha="center", va="bottom", fontsize=9)

Tight layout + save

fig.tight_layout()
fig.savefig("revenue.png", dpi=200)

Worked Example · Polished Bar Chart

12 min

# bar.py — production-quality bar chart
import matplotlib.pyplot as plt

products = ["Pancake", "Cocoa", "Rice", "Tea"]
revenue  = [120, 95, 180, 60]

fig, ax = plt.subplots(figsize=(8, 5))

bars = ax.bar(products, revenue,
              color=["#4a90e2", "#e2884a", "#4ae28d", "#e24a8d"],
              edgecolor="black", linewidth=0.5)

ax.set_title("Revenue by product · May 2026", fontsize=14, pad=10)
ax.set_xlabel("Product")
ax.set_ylabel("Revenue ($)")
ax.set_ylim(0, max(revenue) * 1.15)
ax.grid(axis="y", linestyle="--", alpha=0.4)

for bar in bars:
    h = bar.get_height()
    ax.text(bar.get_x() + bar.get_width() / 2, h + 3,
            f"$ {h}", ha="center", va="bottom", fontsize=10)

fig.tight_layout()
fig.savefig("revenue.png", dpi=200)
plt.show()

Open revenue.png in any image viewer. You get a chart you could paste straight into a school presentation. Five elements made the difference:

Distinct colours per bar.
Annotated values above each bar.
Horizontal grid for easier reading.
Extra y-axis headroom so labels don't collide.
tight_layout prevents the title or labels being clipped.

Basic

5 min

01 🟢 Horizontal bars

Use ax.barh(...) for horizontal bars instead. Useful when names are long.

Challenge 1

4 min

02 🟡 Sort by value

Sort the products so the tallest bar is on the left. Then add a colour scheme that goes dark → light.

Hint

import matplotlib.cm as cm
pairs = sorted(zip(products, revenue), key=lambda p: -p[1])
products, revenue = zip(*pairs)
colors = cm.Blues([1 - i / len(products) for i in range(len(products))])
ax.bar(products, revenue, color=colors)

Challenge 2

4 min

03 🔴 Grouped bars (two series)

Compare last month vs this month: two bars per product, side by side.

Hint

import numpy as np
last = [110, 90, 160, 55]
this = [120, 95, 180, 60]

x = np.arange(len(products))
w = 0.4

fig, ax = plt.subplots()
ax.bar(x - w/2, last, w, label="Apr")
ax.bar(x + w/2, this, w, label="May")
ax.set_xticks(x, products)
ax.legend()

Challenge 3 · Bar From a CSV

8 min

Read clean.csv, compute revenue per product with pandas, draw a sorted bar chart with annotations, save to revenue.png.

Show one possible solution

import pandas as pd, matplotlib.pyplot as plt

df = pd.read_csv("clean.csv")
df["total"] = df["quantity"] * df["price"]
rev = (df.groupby("product")["total"].sum()
        .sort_values(ascending=False))

fig, ax = plt.subplots(figsize=(8, 5))
bars = ax.bar(rev.index, rev.values, color="#4a90e2", edgecolor="black")
ax.set_title("Revenue by product")
ax.set_ylabel("$")
for b in bars:
    ax.text(b.get_x() + b.get_width()/2, b.get_height() + 0.3,
            f"$ {b.get_height():.0f}", ha="center")

fig.tight_layout()
fig.savefig("revenue.png", dpi=200)

Recap

3 min

Bars in 4 lines; presentable bars in 10. Use the OO style (fig, ax = plt.subplots()) — it carries over to subplots tomorrow. Always title + label + save. Annotations beat tooltips for static reports.

Extra Mission

4 min

Draw a polished bar chart of any aggregate from your real CSV. Hit all five rules: title, labels, sorted, annotated, saved as PNG.

# bar.py — production-quality bar chart import matplotlib.pyplot as plt products = ["Pancake", "Cocoa", "Rice", "Tea"] revenue = [120, 95, 180, 60] fig, ax = plt.subplots(figsize=(8, 5)) bars = ax.bar(products, revenue, color=["#4a90e2", "#e2884a", "#4ae28d", "#e24a8d"], edgecolor="black", linewidth=0.5) ax.set_title("Revenue by product · May 2026", fontsize=14, pad=10) ax.set_xlabel("Product") ax.set_ylabel("Revenue ($)") ax.set_ylim(0, max(revenue) * 1.15) ax.grid(axis="y", linestyle="--", alpha=0.4) for bar in bars: h = bar.get_height() ax.text(bar.get_x() + bar.get_width() / 2, h + 3, f"$ {h}", ha="center", va="bottom", fontsize=10) fig.tight_layout() fig.savefig("revenue.png", dpi=200) plt.show()

import matplotlib.cm as cm pairs = sorted(zip(products, revenue), key=lambda p: -p[1]) products, revenue = zip(*pairs) colors = cm.Blues([1 - i / len(products) for i in range(len(products))]) ax.bar(products, revenue, color=colors)

import numpy as np last = [110, 90, 160, 55] this = [120, 95, 180, 60] x = np.arange(len(products)) w = 0.4 fig, ax = plt.subplots() ax.bar(x - w/2, last, w, label="Apr") ax.bar(x + w/2, this, w, label="May") ax.set_xticks(x, products) ax.legend()

import pandas as pd, matplotlib.pyplot as plt df = pd.read_csv("clean.csv") df["total"] = df["quantity"] * df["price"] rev = (df.groupby("product")["total"].sum() .sort_values(ascending=False)) fig, ax = plt.subplots(figsize=(8, 5)) bars = ax.bar(rev.index, rev.values, color="#4a90e2", edgecolor="black") ax.set_title("Revenue by product") ax.set_ylabel("$") for b in bars: ax.text(b.get_x() + b.get_width()/2, b.get_height() + 0.3, f"$ {b.get_height():.0f}", ha="center") fig.tight_layout() fig.savefig("revenue.png", dpi=200)