15.5: Data Visualization

Learning Objectives

By the end of this section you should be able to

• Explain why visualization has an important role in data science.
• Choose appropriate visualization for a given task.
• Use Python visualization libraries to create data visualization.

Why visualization?

Data visualization has a crucial role in data science for understanding the data. Data visualization can be used in all steps of the data science life cycle to facilitate data exploration, identify anomalies, understand relationships and trends, and produce reports. Several visualization types are commonly used:

Visualization type	Description	Benefits/common usage
Bar plot	Rectangular bars	Compare values across different categories.
Line plot	A series of data points connected by line segments	Visualize trends and changes.
Scatter plot	Individual data points representing the relationship between two variables	Identify correlations, clusters, and outliers.
Histogram plot	Rectangular bars representing the distribution of a continuous variable by dividing the variable's range into bins and representing the frequency or count of data within each bin	Summarizing the distribution of the data.
Box plot	Rectangular box with whiskers that summarize the distribution of a continuous variable, including the median, quartiles, and outliers	Summarizing the distribution of the data and comparing different variables.

Table 15.7 Common visualization types.

Checkpoint: Visualization types

Access multimedia content

Concepts in Practice: Comparing visualization methods

1.

Which of the following plot types is best suited for comparing the distribution of a continuous variable?

1. scatter plot

histogram plot

line plot

2.

Which of the following plot types is best suited for visualizing outliers and quartiles of a continuous variable?

1. histogram plot
2. bar plot
3. box plot

3.

Which of the following plot types is effective for displaying trends and changes over time?

1. line plot
2. bar plot
3. histogram plot

Data visualization tools

Many Python data visualization libraries exist that offer a range of capabilities and features to create different plot types. Some of the most commonly used frameworks are Matplotlib, Plotly, and Seaborn. Here, some useful functionalities of Matplotlib are summarized.

Plot type	Method
Bar plot	The plt.bar(x, height) function takes in two inputs, x and height, and plots bars for each x value with the height given in the height variable.
Example	Output
import matplotlib.pyplot as plt # Data categories = ["Course A", "Course B", "Course C"] values = [25, 40, 30] # Create the bar chart fig = plt.bar(categories, values) # Customize the chart plt.title("Number of students in each course') plt.xlabel("Courses") plt.ylabel("Number of students") # Display the chart plt.show()

Table 15.8 Matplotlib functionalities. Bar plot.

Plot type	Method
Line plot	The plt.plot(x, y) function takes in two inputs, x and y, and plots lines connecting pairs of (x, y) values.
Example	Output
import matplotlib.pyplot as plt # Data month = ["Jan", "Feb", "Mar", "Apr", "May"] inflation = [6.41, 6.04, 4.99, 4.93, 4.05] # Create the line chart plt.plot(month, inflation, marker="o", linestyle="-", color="blue") # Customize the chart plt.title("Inflation trend in 2023") plt.xlabel("Month") plt.ylabel("Inflation") # Display the chart plt.show()

Table 15.9 Matplotlib functionalities. Line plot.

Plot type	Method
Scatter plot	The plt.scatter(x, y) function takes in two inputs, x and y, and plots points representing (x, y) pairs.
Example	Output
import matplotlib.pyplot as plt # Data x = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10] y = [10, 8, 6, 4, 2, 5, 7, 9, 3, 1] # Create the scatter plot plt.scatter(x, y, marker="o", color="blue") # Customize the chart plt.title("Scatter Plot Example") plt.xlabel("X") plt.ylabel("Y") # Display the chart plt.show()

Table 15.10 Matplotlib functionalities. Scatter plot.

Plot type	Method
Histogram plot	The plt.hist(x) function takes in one input, x, and plots a histogram of values in x to show distribution or trend.
Example	Output
import matplotlib.pyplot as plt import numpy as np # Data: random 1000 samples data = np.random.randn(1000) # Create the histogram plt.hist(data, bins=30, edgecolor="black") # Customize the chart plt.title("Histogram of random values") plt.xlabel("Values") plt.ylabel("Frequency") # Display the chart plt.show()

Table 15.11 Matplotlib functionalities. Histogram plot.

Plot type	Method
Box plot	The plt.boxplot(x) function takes in one input, x, and represents minimum, maximum, first, second, and third quartiles, as well as outliers in x.
Example	Output
import matplotlib.pyplot as plt import numpy as np # Data: random 100 samples data = [np.random.normal(0, 5, 100)] # Create the box plot plt.boxplot(data) # Customize the chart plt.title("Box Plot of random values") plt.xlabel("Data Distribution") plt.ylabel("Values") # Display the chart plt.show()

Table 15.12 Matplotlib functionalities. Box plot.

Concepts in Practice: Matplotlib methods

4.

Given the following code, which of the function calls is appropriate in showing association between x and y?

import matplotlib.pyplot as plt

# Data
x = [1, 2, 3, 4, 5]
y = [10, 15, 12, 18, 20]

1. plt.boxplot(x)

plt.hist(y)

plt.scatter(x, y)

5.

What is being plotted using the code below?

import matplotlib.pyplot as plt

# Data
categories = ['A', 'B', 'C', 'D']
values = [10, 15, 12, 18]
plt.bar(categories, values)

1. a histogram plot
2. a bar plot
3. a line plot

6.

Which library in Python is more commonly used for creating interactive visualizations?

1. Matplotlib
2. Plotly
3. Pandas

Exploring further

Please refer to the following user guide for more information about the Matplotlib, Plotly, and Seaborn libraries.

• Matplotlib User Guide
• Plotly User Guide
• Seaborn User Guide

Programming practice with Google

Use the Google Colaboratory document below to practice a visualization task on a given dataset.

Google Colaboratory document