Bar Charts

What Is a Bar Chart?

A bar chart represents categories using rectangular bars. The length or height of each bar represents its value.

Laptop  → 500
Phone   → 800
Tablet  → 300
Monitor → 600

When Should You Use a Bar Chart?

Use a bar chart when you want to compare separate categories.

• Sales by product
• Revenue by country
• Number of students by class
• Orders by category
• Website visitors by page

Different categories
        ↓
Compare their values
        ↓
Use a bar chart

Bar Chart vs Line Chart

This distinction is important. Don't choose a chart just because you know its syntax.

January
February
March
April
May

India
USA
Germany
Japan

Import Matplotlib

We use Matplotlib's pyplot module to create the chart.

import matplotlib.pyplot as plt

The plt name is simply a shorter way to access matplotlib.pyplot.

Create Your First Bar Chart

Matplotlib uses plt.bar() to create a vertical bar chart.

import matplotlib.pyplot as plt

products = [
    "Laptop",
    "Phone",
    "Tablet",
    "Monitor"
]

sales = [
    500,
    800,
    300,
    600
]

plt.bar(
    products,
    sales
)

plt.show()

The first argument contains the categories and the second argument contains their values.

plt.bar(
    categories,
    values
)

Understand How the Bars Work

Matplotlib matches each category with its corresponding value.

products = ["Laptop", "Phone", "Tablet"]

sales = [500, 800, 300]

This creates:

Laptop → 500
Phone  → 800
Tablet → 300

The value controls the height of each bar.

800 |       █
700 |       █
600 |       █
500 |   █   █
400 |   █   █
300 |   █   █   █
200 |   █   █   █
100 |   █   █   █
    +---------------
      Laptop Phone Tablet

Add a Title and Labels

A good chart should clearly explain what the data represents.

import matplotlib.pyplot as plt

products = [
    "Laptop",
    "Phone",
    "Tablet",
    "Monitor"
]

sales = [
    500,
    800,
    300,
    600
]

plt.bar(
    products,
    sales
)

plt.title("Product Sales")

plt.xlabel("Product")

plt.ylabel("Sales")

plt.show()

Now the chart tells us:

Title
  ↓
Product Sales

X-axis
  ↓
Product

Y-axis
  ↓
Sales

Display Values on the Bars

Sometimes you want the exact value to appear above each bar.

Matplotlib can do this using bar_label().

import matplotlib.pyplot as plt

products = [
    "Laptop",
    "Phone",
    "Tablet",
    "Monitor"
]

sales = [
    500,
    800,
    300,
    600
]

bars = plt.bar(
    products,
    sales
)

plt.bar_label(bars)

plt.title("Product Sales")

plt.xlabel("Product")
plt.ylabel("Sales")

plt.show()

The important part is:

bars = plt.bar(...)

plt.bar_label(bars)

First we store the bars in the bars variable. Then bar_label() adds the values to them.

Horizontal Bar Charts

You can also display bars horizontally using plt.barh().

import matplotlib.pyplot as plt

products = [
    "Laptop",
    "Phone",
    "Tablet",
    "Monitor"
]

sales = [
    500,
    800,
    300,
    600
]

plt.barh(
    products,
    sales
)

plt.title("Product Sales")

plt.xlabel("Sales")
plt.ylabel("Product")

plt.show()

The difference is simple:

plt.bar()
    ↓
Vertical bars

plt.barh()
    ↓
Horizontal bars

Sorting Bar Chart Data

When comparing categories, sorting the values can make the comparison easier.

products = [
    "Laptop",
    "Phone",
    "Tablet",
    "Monitor"
]

sales = [
    500,
    800,
    300,
    600
]

The values are:

Phone  → 800
Monitor → 600
Laptop → 500
Tablet → 300

A sorted bar chart makes the ranking immediately visible.

Phone   █████████ 800
Monitor ██████    600
Laptop  █████     500
Tablet  ███       300

This is particularly useful for rankings and top-performing categories.

Bar Charts With Pandas

You can combine Pandas and Matplotlib just like you did with line charts.

import pandas as pd
import matplotlib.pyplot as plt

sales = pd.DataFrame({
    "Product": [
        "Laptop",
        "Phone",
        "Tablet",
        "Monitor"
    ],
    "Sales": [
        500,
        800,
        300,
        600
    ]
})

plt.bar(
    sales["Product"],
    sales["Sales"]
)

plt.title("Product Sales")

plt.xlabel("Product")
plt.ylabel("Sales")

plt.show()

The responsibilities remain separate:

Pandas
  ↓
Store / clean / analyze data
  ↓
Matplotlib
  ↓
Visualize data

Real-World Example: Sales by Country

Imagine an online business wants to compare sales from different countries.

import matplotlib.pyplot as plt

countries = [
    "India",
    "USA",
    "Germany",
    "Japan"
]

sales = [
    1200,
    2500,
    1800,
    1400
]

bars = plt.bar(
    countries,
    sales
)

plt.bar_label(bars)

plt.title("Sales by Country")

plt.xlabel("Country")
plt.ylabel("Sales")

plt.show()

The purpose here is comparison, not showing a continuous trend. Therefore, a bar chart makes more sense than a line chart.

Bar Charts in AI

Bar charts are also useful in AI and Machine Learning. For example, you can compare the accuracy of different models.

import matplotlib.pyplot as plt

models = [
    "Logistic Regression",
    "Decision Tree",
    "Random Forest",
    "Neural Network"
]

accuracy = [
    82,
    85,
    91,
    94
]

bars = plt.bar(
    models,
    accuracy
)

plt.bar_label(bars)

plt.title("Model Accuracy")

plt.xlabel("Model")

plt.ylabel("Accuracy (%)")

plt.show()

Now you can immediately compare the performance of the models.

Logistic Regression → 82%
Decision Tree       → 85%
Random Forest       → 91%
Neural Network      → 94%

Complete Bar Chart Example

Here is a complete example combining the important concepts.

import matplotlib.pyplot as plt

models = [
    "Logistic Regression",
    "Decision Tree",
    "Random Forest",
    "Neural Network"
]

accuracy = [
    82,
    85,
    91,
    94
]

bars = plt.bar(
    models,
    accuracy
)

plt.bar_label(bars)

plt.title("Model Accuracy")

plt.xlabel("Model")

plt.ylabel("Accuracy (%)")

plt.xticks(rotation=15)

plt.show()

The workflow is:

Create categories
        ↓
Create values
        ↓
plt.bar()
        ↓
Add title
        ↓
Add axis labels
        ↓
Add values if needed
        ↓
Display chart

The Most Important Rule

The key question is not "How do I create a bar chart?" It is "Is a bar chart the correct visualization for this data?"

Product A → 500
Product B → 800
Product C → 300
Product D → 600

January  → 500
February → 600
March    → 700
April    → 850