Bar Chart and Horizontal Bar

Bar charts compare a value across discrete categories — each category gets one bar, and the bar’s height (or length, for horizontal bars) shows its value. As McKinney notes, plot.bar() and plot.barh() even work directly with a labeled Series or DataFrame: the index becomes the category ticks automatically.

diagram Vertical vs horizontal bars mermaid

  flowchart LR
  A["Categories + values"] --> B{"Long labels?"}
  B -->|No| C["plt.bar() - vertical bars"]
  B -->|Yes| D["plt.barh() - horizontal bars"]
  C --> E["Sort by value for easy scanning"]
  D --> E

Both encode a value as a bar's length; horizontal bars just swap which axis carries the category.

Basic bar chart

Bar

import matplotlib.pyplot as plt
 
cities = ["Pune", "Delhi", "Mumbai", "Bengaluru"]
sales = [120, 180, 90, 160]
 
plt.figure(figsize=(7, 4))
plt.bar(cities, sales)
plt.title("Sales by city")
plt.xlabel("City")
plt.ylabel("Sales")
plt.xticks(rotation=20)
plt.tight_layout()
plt.show()

Sorted bars

Sorted bars

import matplotlib.pyplot as plt
 
data = {"Pune": 120, "Delhi": 180, "Mumbai": 90, "Bengaluru": 160}
items = sorted(data.items(), key=lambda x: x[1], reverse=True)
labels = [k for k, _ in items]
values = [v for _, v in items]
 
plt.figure(figsize=(7, 4))
plt.bar(labels, values)
plt.title("Sales by city (sorted)")
plt.tight_layout()
plt.show()

Horizontal bars

Horizontal bar

import matplotlib.pyplot as plt
 
labels = ["Very long category A", "Long B", "C"]
values = [12, 19, 7]
 
plt.figure(figsize=(7, 4))
plt.barh(labels, values)
plt.title("Horizontal bar chart")
plt.xlabel("Value")
plt.tight_layout()
plt.show()

Visualize it

Watch each category’s bar grow to its value — the tallest bar is the easiest to spot because bars share the same zero baseline:

sketch A bar chart compares categories p5.js Pause Restart

Each bar's height is proportional to its value, starting from a shared zero baseline.

From the book

McKinney’s tip: a stacked bar chart (stacked=True) is a quick way to see how a whole splits across sub-categories — for example, party size share by day of week. But once you have more than three or four stacked segments, a stacked bar becomes as hard to read as a pie chart. Consider small multiples (separate subplots) instead.

Tip

Start y-axis at 0 for bar charts to avoid misleading comparisons.

Next

Continue to Histogram to visualize the distribution of a single numeric column.

🧪 Try It Yourself

Exercise 1 – Draw vertical bars

pch.exerciseTag pch.exerciseTitle

Exercise 2 – Horizontal bars for long labels

pch.exerciseTag pch.exerciseTitle

Exercise 3 – Sort before plotting

pch.exerciseTag pch.exerciseTitle