Data Visualization With Python PDF

Contents
1. Preface
1. About the Book
1. About the Author

2. Objectives
3. Audience
4. Approach
5. Minimum Hardware Requirements
6. Software Requirements
7. Conventions
8. Installation and Setup
9. Working with JupyterLab and Jupyter Notebook
10. Importing Python Libraries
11. Installing the Code Bundle
12. Additional Resources
2. Chapter 1
3. The Importance of Data Visualization and Data Exploration
1. Introduction
1. Introduction to Data Visualization

2. The Importance of Data Visualization
3. Data Wrangling
4. Tools and Libraries for Visualization
2. Overview of Statistics
1. Measures of Central Tendency

2. Measures of Dispersion
3. Correlation
4. Types of Data
5. Summary Statistics
3. NumPy
1. Exercise 1: Loading a Sample Dataset and

Calculating the Mean
2. Activity 1: Using NumPy to Compute the Mean,
Median, Variance, and Standard Deviation for
the Given Numbers
3. Basic NumPy Operations
4. Activity 2: Indexing, Slicing, Splitting, and
Iterating
5. Advanced NumPy Operations
6. Activity 3: Filtering, Sorting, Combining, and
Reshaping
4. pandas
1. Advantages of pandas over NumPy

2. Disadvantages of pandas
3. Exercise 2: Loading a Sample Dataset and
Calculating the Mean
4. Activity 4: Using pandas to Compute the Mean,
Median, and Variance for the Given Numbers
5. Basic Operations of pandas
6. Series
7. Activity 5: Indexing, Slicing, and Iterating using
pandas
8. Advanced pandas Operations
9. Activity 6: Filtering, Sorting, and Reshaping
5. Summary
4. Chapter 2
5. All You Need to Know About Plots
1. Introduction
2. Comparison Plots
1. Line Chart
2. Bar Chart
3. Radar Chart
4. Activity 7: Employee Skill Comparison
3. Relation Plots
1. Scatter Plot
2. Bubble Plot
3. Correlogram
4. Heatmap
5. Activity 8: Road Accidents Occurring over Two
Decades
4. Composition Plots
1. Pie Chart
2. Stacked Bar Chart
3. Stacked Area Chart
4. Activity 9: Smartphone Sales Units
5. Venn Diagram
5. Distribution Plots
1. Histogram
2. Density Plot
3. Box Plot
4. Violin Plot
5. Activity 10: Frequency of Trains during
Different Time Intervals
6. Geo Plots
1. Dot Map
2. Choropleth Map
3. Connection Map
7. What Makes a Good Visualization?
1. Activity 11: Identifying the Ideal Visualization
8. Summary
6. Chapter 3
7. A Deep Dive into Matplotlib
1. Introduction
2. Overview of Plots in Matplotlib
3. Pyplot Basics
1. Creating Figures
2. Closing Figures
3. Format Strings
4. Plotting
5. Plotting Using pandas DataFrames
6. Displaying Figures
7. Saving Figures
8. Exercise 3: Creating a Simple Visualization
4. Basic Text and Legend Functions
1. Labels
2. Titles
3. Text
4. Annotations
5. Legends
6. Activity 12: Visualizing Stock Trends by Using a
Line Plot
5. Basic Plots
1. Bar Chart
2. Activity 13: Creating a Bar Plot for Movie
Comparison
3. Pie Chart
4. Exercise 4: Creating a Pie Chart for Water Usage
5. Stacked Bar Chart
6. Activity 14: Creating a Stacked Bar Plot to
Visualize Restaurant Performance
7. Stacked Area Chart
8. Activity 15: Comparing Smartphone Sales Units
Using a Stacked Area Chart
9. Histogram
10. Box Plot
11. Activity 16: Using a Histogram and a Box Plot
to Visualize the Intelligence Quotient
12. Scatter Plot
13. Activity 17: Using a Scatter Plot to Visualize
Correlation Between Various Animals
14. Bubble Plot
6. Layouts
1. Subplots
2. Tight Layout
3. Radar Charts
4. Exercise 5: Working on Radar Charts
5. GridSpec
6. Activity 18: Creating Scatter Plot with Marginal
Histograms
7. Images
1. Basic Image Operations

2. Activity 19: Plotting Multiple Images in a Grid
8. Writing Mathematical Expressions

9. Summary
8. Chapter 4
9. Simplifying Visualizations Using Seaborn
1. Introduction
1. Advantages of Seaborn
2. Controlling Figure Aesthetics
1. Seaborn Figure Styles

2. Removing Axes Spines
3. Contexts
4. Activity 20: Comparing IQ Scores for Different
Test Groups by Using a Box Plot
3. Color Palettes
1. Categorical Color Palettes

2. Sequential Color Palettes
3. Diverging Color Palettes
4. Activity 21: Using Heatmaps to Find Patterns in
Flight Passengers' Data
4. Interesting Plots in Seaborn
1. Bar Plots
2. Activity 22: Movie Comparison Revisited
3. Kernel Density Estimation
4. Plotting Bivariate Distributions
5. Visualizing Pairwise Relationships
6. Violin Plots
Test Groups by Using a Violin Plot
5. Multi-Plots in Seaborn
1. FacetGrid
2. Activity 24: Top 30 YouTube Channels
6. Regression Plots
1. Activity 25: Linear Regression
7. Squarify
1. Activity 26: Water Usage Revisited
8. Summary
10. Chapter 5
11. Plotting Geospatial Data
1. Introduction
1. The Design Principles of Geoplotlib

2. Geospatial Visualizations
3. Exercise 6: Visualizing Simple Geospatial Data
4. Activity 27: Plotting Geospatial Data on a Map
5. Exercise 7: Choropleth Plot with GeoJSON Data
2. Tile Providers
1. Exercise 8: Visually Comparing Different Tile

Providers
3. Custom Layers
1. Activity 28: Working with Custom Layers
4. Summary
12. Chapter 6
13. Making Things Interactive with Bokeh
1. Introduction
1. Concepts of Bokeh
2. Interfaces in Bokeh
3. Output
4. Bokeh Server
5. Presentation
6. Integrating
7. Exercise 9: Plotting with Bokeh
8. Exercise 10: Comparing the Plotting and Models
Interfaces
2. Adding Widgets
1. Exercise 11: Basic Interactivity Widgets

2. Activity 29: Extending Plots with Widgets
3. Summary
14. Chapter 7
15. Combining What We Have Learned
1. Introduction
1. Activity 30: Implementing Matplotlib and

Seaborn on New York City Database
2. Bokeh
3. Activity 31: Visualizing Bokeh Stock Prices
4. Geoplotlib
5. Activity 32: Analyzing Airbnb Data with
geoplotlib
2. Summary
16. Appendix
1. Chapter 1: The Importance of Data Visualization and Data

Exploration
1. Activity 1: Using NumPy to Compute the Mean,

Median, Variance, and Standard Deviation for
the Given Numbers
2. Activity 2: Indexing, Slicing, Splitting, and
Iterating
3. Activity 3: Filtering, Sorting, Combining, and
Reshaping
4. Activity 4: Using pandas to Compute the Mean,
Median, and Variance for the Given Numbers
5. Activity 5: Indexing, Slicing, and Iterating Using
pandas
6. Activity 6: Filtering, Sorting, and Reshaping
2. Chapter 2: All You Need to Know about Plots
1. Activity 7: Employee Skill Comparison

2. Activity 8: Road Accidents Occurring over Two
Decades
3. Activity 9: Smartphone Sales Units
4. Activity 10: Frequency of Trains during
Different Time Intervals
5. Activity 11: Identifying the Ideal Visualization
3. Chapter 3: A Deep Dive into Matplotlib
1. Activity 12: Visualizing Stock Trends by Using a

Line Plot
2. Activity 13: Creating a Bar Plot for Movie
Comparison
3. Activity 14: Creating a Stacked Bar Plot to
Visualize Restaurant Performance
4. Activity 15: Comparing Smartphone Sales Units
Using a Stacked Area Chart
5. Activity 16: Using a Histogram and a Box Plot
to Visualize the Intelligence Quotient
6. Activity 17: Using a Scatter Plot to Visualize
Correlation between Various Animals
7. Activity 18: Creating a Scatter Plot with
Marginal Histograms
8. Activity 19: Plotting Multiple Images in a Grid
4. Chapter 4: Simplifying Visualizations Using Seaborn

Test Groups by Using a Box Plot
2. Activity 21: Using Heatmaps to Find Patterns in
Flight Passengers' Data
3. Activity 22: Movie Comparison Revisited
Test Groups by Using a Violin Plot
5. Activity 24: Top 30 YouTube Channels
6. Activity 25: Linear Regression
7. Activity 26: Water Usage Revisited
5. Chapter 5: Plotting Geospatial Data
1. Activity 27: Plotting Geospatial Data on a Map

2. Activity 28: Working with Custom Layers
6. Chapter 6: Making Things Interactive with Bokeh
1. Activity 29: Extending Plots with Widgets
7. Chapter 7: Combining What We Have Learned
1. Activity 30: Implementing Matplotlib and

Seaborn on New York City Database
2. Activity 31: Bokeh Stock Prices Visualization
3. Activity 32: Analyzing Airbnb Data with
Geoplotlib
Landmarks
1. Cover
2. Table of Contents
DATA VISUALIZATION
WITH PYTHON
Copyright © 2019 Packt Publishing
All rights reserved. No part of this book may be reproduced,

stored in a retrieval system, or transmitted in any form or by
any means, without the prior written permission of the
publisher, except in the case of brief quotations embedded in
critical articles or reviews.
Every effort has been made in the preparation of this book to

ensure the accuracy of the information presented. However, the
information contained in this book is sold without warranty,
either express or implied. Neither the author, nor Packt
Publishing, and its dealers and distributors will be held liable
for any damages caused or alleged to be caused directly or
indirectly by this book.
Packt Publishing has endeavored to provide trademark

information about all of the companies and products mentioned
in this book by the appropriate use of capitals. However, Packt
Publishing cannot guarantee the accuracy of this information.
Author: Mario Döbler and Tim Größmann
Reviewer: José España
Managing Editor: Aditya Shah and Bhavesh Bangera
Acquisitions Editor: Kunal Sawant
Production Editor: Shantanu Zagade
Editorial Board: David Barnes, Ewan Buckingham, Shivangi

Chatterji, Simon Cox, Manasa Kumar, Alex Mazonowicz,
Douglas Paterson, Dominic Pereira, Shiny Poojary, Saman
Siddiqui, Erol Staveley, Ankita Thakur and Mohita Vyas
First Published: February 2019
Production Reference: 1280219
ISBN: 978-1-78995-646-7
Table of Contents
Preface
The Importance of
Data Visualization and
Data Exploration
INTRODUCTION
INTRODUCTION TO DATA
VISUALIZATION
THE IMPORTANCE OF
DATA VISUALIZATION
DATA WRANGLING
TOOLS AND LIBRARIES
FOR VISUALIZATION
OVERVIEW OF
STATISTICS
MEASURES OF CENTRAL
TENDENCY
MEASURES OF
DISPERSION
CORRELATION
TYPES OF DATA
SUMMARY STATISTICS
NUMPY
EXERCISE 1: LOADING A
SAMPLE DATASET AND
CALCULATING THE MEAN
ACTIVITY 1: USING
NUMPY TO COMPUTE
THE MEAN, MEDIAN,
VARIANCE, AND
STANDARD DEVIATION
FOR THE GIVEN
NUMBERS
BASIC NUMPY
OPERATIONS
ACTIVITY 2: INDEXING,
SLICING, SPLITTING, AND
ITERATING
ADVANCED NUMPY
OPERATIONS
ACTIVITY 3: FILTERING,
SORTING, COMBINING,
AND RESHAPING
PANDAS
ADVANTAGES OF
PANDAS OVER NUMPY
DISADVANTAGES OF
PANDAS
SAMPLE DATASET AND
ACTIVITY 4: USING
PANDAS TO COMPUTE
THE MEAN, MEDIAN, AND
VARIANCE FOR THE
GIVEN NUMBERS
BASIC OPERATIONS OF
PANDAS
SERIES
SLICING, AND ITERATING
USING PANDAS
ADVANCED PANDAS
OPERATIONS
SORTING, AND
RESHAPING
SUMMARY
All You Need to Know

About Plots
INTRODUCTION
COMPARISON PLOTS
LINE CHART
BAR CHART
RADAR CHART
ACTIVITY 7: EMPLOYEE
SKILL COMPARISON
RELATION PLOTS
SCATTER PLOT
BUBBLE PLOT
CORRELOGRAM
HEATMAP
ACTIVITY 8: ROAD
ACCIDENTS OCCURRING
OVER TWO DECADES
COMPOSITION PLOTS
PIE CHART
STACKED BAR CHART

STACKED AREA CHART
ACTIVITY 9:
SMARTPHONE SALES
UNITS
VENN DIAGRAM
DISTRIBUTION PLOTS
HISTOGRAM
DENSITY PLOT
BOX PLOT
VIOLIN PLOT
ACTIVITY 10: FREQUENCY
OF TRAINS DURING
DIFFERENT TIME
INTERVALS
GEO PLOTS
DOT MAP
CHOROPLETH MAP
CONNECTION MAP
WHAT MAKES A GOOD

VISUALIZATION?
ACTIVITY 11: IDENTIFYING
THE IDEAL
VISUALIZATION
SUMMARY
A Deep Dive into

Matplotlib
INTRODUCTION
OVERVIEW OF PLOTS IN
MATPLOTLIB
PYPLOT BASICS
CREATING FIGURES
CLOSING FIGURES
FORMAT STRINGS
PLOTTING
PLOTTING USING
PANDAS DATAFRAMES
DISPLAYING FIGURES
SAVING FIGURES
EXERCISE 3: CREATING A
SIMPLE VISUALIZATION
BASIC TEXT AND LEGEND
FUNCTIONS
LABELS
TITLES
TEXT
ANNOTATIONS
LEGENDS
ACTIVITY 12: VISUALIZING

STOCK TRENDS BY
USING A LINE PLOT
BASIC PLOTS
BAR CHART
ACTIVITY 13: CREATING A

BAR PLOT FOR MOVIE
COMPARISON
PIE CHART
PIE CHART FOR WATER
USAGE
STACKED BAR CHART

STACKED BAR PLOT TO
VISUALIZE RESTAURANT
PERFORMANCE
STACKED AREA CHART
ACTIVITY 15: COMPARING

SMARTPHONE SALES
UNITS USING A STACKED
AREA CHART
HISTOGRAM
BOX PLOT
ACTIVITY 16: USING A
HISTOGRAM AND A BOX
PLOT TO VISUALIZE THE
INTELLIGENCE QUOTIENT
SCATTER PLOT

SCATTER PLOT TO
VISUALIZE CORRELATION
BETWEEN VARIOUS
ANIMALS
BUBBLE PLOT
LAYOUTS
SUBPLOTS
TIGHT LAYOUT
RADAR CHARTS
EXERCISE 5: WORKING
ON RADAR CHARTS
GRIDSPEC
ACTIVITY 18: CREATING

SCATTER PLOT WITH
MARGINAL HISTOGRAMS
IMAGES
BASIC IMAGE
OPERATIONS
ACTIVITY 19: PLOTTING

MULTIPLE IMAGES IN A
GRID
WRITING MATHEMATICAL
EXPRESSIONS
SUMMARY
Simplifying
Visualizations Using
Seaborn
INTRODUCTION
ADVANTAGES OF
SEABORN
CONTROLLING FIGURE
AESTHETICS
SEABORN FIGURE
STYLES
REMOVING AXES SPINES

CONTEXTS

IQ SCORES FOR
DIFFERENT TEST
GROUPS BY USING A BOX
PLOT
COLOR PALETTES
CATEGORICAL COLOR
PALETTES
SEQUENTIAL COLOR
PALETTES
DIVERGING COLOR
PALETTES
ACTIVITY 21: USING

HEATMAPS TO FIND
PATTERNS IN FLIGHT
PASSENGERS' DATA
INTERESTING PLOTS IN
SEABORN
BAR PLOTS
ACTIVITY 22: MOVIE

COMPARISON REVISITED
KERNEL DENSITY
ESTIMATION
PLOTTING BIVARIATE
DISTRIBUTIONS
VISUALIZING PAIRWISE
RELATIONSHIPS
VIOLIN PLOTS

IQ SCORES FOR
DIFFERENT TEST
GROUPS BY USING A
VIOLIN PLOT
MULTI-PLOTS IN
SEABORN
FACETGRID
ACTIVITY 24: TOP 30

YOUTUBE CHANNELS
REGRESSION PLOTS
ACTIVITY 25: LINEAR

REGRESSION
SQUARIFY
ACTIVITY 26: WATER

USAGE REVISITED
SUMMARY
Plotting Geospatial
Data
INTRODUCTION
THE DESIGN PRINCIPLES
OF GEOPLOTLIB
GEOSPATIAL
VISUALIZATIONS
EXERCISE 6: VISUALIZING
SIMPLE GEOSPATIAL
DATA

GEOSPATIAL DATA ON A
MAP
EXERCISE 7:
CHOROPLETH PLOT WITH
GEOJSON DATA
TILE PROVIDERS
EXERCISE 8: VISUALLY
COMPARING DIFFERENT
TILE PROVIDERS
CUSTOM LAYERS
ACTIVITY 28: WORKING

WITH CUSTOM LAYERS
SUMMARY
Making Things
Interactive with Bokeh
INTRODUCTION
CONCEPTS OF BOKEH
INTERFACES IN BOKEH
OUTPUT
BOKEH SERVER
PRESENTATION
INTEGRATING
EXERCISE 9: PLOTTING
WITH BOKEH
EXERCISE 10:
COMPARING THE
PLOTTING AND MODELS
INTERFACES
ADDING WIDGETS
EXERCISE 11: BASIC

INTERACTIVITY WIDGETS
ACTIVITY 29: EXTENDING

PLOTS WITH WIDGETS
SUMMARY
Combining What We
Have Learned
INTRODUCTION
ACTIVITY 30:
IMPLEMENTING
MATPLOTLIB AND
SEABORN ON NEW YORK
CITY DATABASE
BOKEH

BOKEH STOCK PRICES
GEOPLOTLIB
ACTIVITY 32: ANALYZING

AIRBNB DATA WITH
GEOPLOTLIB
SUMMARY
Appendix
Preface
About
This section briefly introduces the author, the coverage of this
book, the technical skills you'll need to get started, and the
hardware and software requirements required to complete all of
the included activities and exercises.
About the Book

You'll begin Data Visualization with Python with an
introduction to data visualization and its importance. Then,
you'll learn about statistics by computing mean, median, and
variance for some numbers, and observing the difference in
their values. You'll also learn about key NumPy and Pandas
techniques, such as indexing, slicing, iterating, filtering, and
grouping. Next, you'll study different types of visualizations,
compare them, and find out how to select a particular type of
visualization using this comparison. You'll explore different
plots, including custom creations.
After you get a hang of the various visualization libraries,

you'll learn to work with Matplotlib and Seaborn to simplify
the process of creating visualizations. You'll also be introduced
to advanced visualization techniques, such as geoplots and
interactive plots. You'll learn how to make sense of geospatial
data, create interactive visualizations that can be integrated into
any webpage, and take any dataset to build beautiful and
insightful visualizations. You'll study how to plot geospatial
data on a map using Choropleth plot, and study the basics of
Bokeh, extending plots by adding widgets and animating the
display of information.
This book ends with an interesting activity in which you will

be given a new dataset and you must apply all that you've
learned to create an insightful capstone visualization.
ABOUT THE AUTHOR

Mario Döbler is a Ph.D. student with focus in deep learning at
the University of Stuttgart. He previously interned at the Bosch
Center for Artificial Intelligence in Silicon Valley in the field
of deep learning, using state-of-the-art algorithms to develop
cutting-edge products. In his master thesis, he dedicated
himself to apply deep learning to medical data to drive medical
applications.
Tim Großmann is a CS student with an interest in diverse

topics, ranging from AI to IoT. He previously worked at the
Bosch Center for Artificial Intelligence in Silicon Valley, in the
field of big data engineering. He's highly involved in different
open source projects and actively speaks at meetups and
conferences about his projects and experiences.
OBJECTIVES
Get an overview of various plots
and their best use cases
Work with different plotting

libraries and get to know their
strengths and weaknesses
Learn how to create insightful

visualizations
Understand what makes a good

visualization
Improve your Python data

wrangling skills
Work with real-world data
Learn industry standard tools
Develop your general

understanding of data formats and
representations
AUDIENCE
This book is aimed at developers or scientists who want to get
into data science or want to use data visualizations to enrich
their personal and professional projects. Prior experience in
data analytics and visualization is not needed; however, some
knowledge of Python and high-school level math is
recommended. Even though this is a beginner-level book on
data visualization, more experienced students will benefit from
improving their Python skills by working with real-world data.
APPROACH
This book thoroughly explains the technology in easy-to-
understand language while perfectly balancing theory and
exercises. Each chapter is designed to build on the learning
from the previous chapter. The book also contains multiple
activities that use real-life business scenarios for you to
practice and apply your new skills in a highly relevant context.
MINIMUM HARDWARE
REQUIREMENTS
For the optimal student experience, we recommend the
following hardware configuration:
OS: Windows 7 SP1 32/64-bit,
Windows 8.1 32/64-bit or Windows
10 32/64-bit, Ubuntu 14.04 or later,
or macOS Sierra or later
Processor: Dual Core or better
Memory: 4GB RAM
Storage: 10 GB available space
SOFTWARE
REQUIREMENTS
You'll also need the following software installed in advance:
Browser: Google Chrome or

Mozilla Firefox
Conda
JupyterLab and Jupyter Notebook
Sublime Text (latest version), Atom

IDE (latest version), or other
similar text editor applications
Python 3
The following Python libraries

installed: NumPy, pandas,
Matplotlib, seaborn, geoplotlib,
Bokeh, and squarify
CONVENTIONS
Code words in text, database table names, folder names,
filenames, file extensions, pathnames, dummy URLs, user
input, and Twitter handles are shown as follows: "axis=0 is
horizontal and axis=1 is vertical, so if we want to have the
result for each row, we need to choose axis=1".
A block of code is set as follows:
# indexing the first value of the second

row (1st row, 1st value)
first_val_first_row = dataset[0][0]
np.mean(first_val_first_row)
New terms and important words are shown in bold:
"To draw conclusions from visualized data, we need to handle

our data and transform it into the best possible representation.
This is where data wrangling is used."
INSTALLATION AND
SETUP
Before you start this book, we'll install Python 3.6, pip, and the
other libraries used throughout this book. You will find the
steps to install them here.
Installing Python
Install Python 3.6 following the instructions in this link:

https://realpython.com/installing-python/.
Installing pip
1. To install pip, go to the following

link and download the get-
pip.py file:
https://pip.pypa.io/en/stable/installi
ng/.
2. Then, use the following command

to install it:
python get-pip.py
You might need to use the

python3 get-pip.py
command, due to previous versions
of Python on your computer that
already use the python command.
Installing libraries
Using the pip command, install the following libraries:
python -m pip install --user numpy

matplotlib jupyterlab pandas squarify
bokeh geoplotlib seaborn
WORKING WITH
JUPYTERLAB AND
JUPYTER NOTEBOOK
You'll be working on different exercises and activities in
JupyterLab. These exercises and activities can be downloaded
from the associated GitHub repository.
Download the repository from here:

https://github.com/TrainingByPackt/Data-Visualization-with-
Python.
You can either download it using GitHub or as a zipped folder
by clicking on the green Clone or download button on the
upper-right side.
In order to open Jupyter Notebooks, you have to traverse into

the directory with your terminal. To do that, type:
cd Data-Visualization-with-Python/<your
current chapter>.
For example:
cd Data-Visualization-with-
Python/chapter01/
To complete the process, perform the following steps:
1. To reach each activity and exercise,

you have to use cd once more to go
into each folder, like so:
cd Activity01
2. Once you are in the folder of your

choice, simply call jupyter-lab
to start up JupyterLab. Similarly,
for Jupyter Notebook, call
jupyter notebook.
IMPORTING PYTHON
LIBRARIES
Every exercise and activity in this book will make use of
various libraries. Importing libraries into Python is very simple
and here's how we do it:
1. To import libraries, such as NumPy
and pandas, we have to run the
following code. This will import
the whole numpy library into our
current file:
import numpy # import

numpy
2. In the first cells of the exercises and

activities of this bookware, you will
see the following code. We can use
np instead of numpy in our code to
call methods from numpy:
import numpy as np #
import numpy and
assign alias np
3. In later chapters, partial imports

will be present, as shown in the
following code. This only loads the
mean method from the library:
from numpy import mean

# only import the mean
method of numpy
INSTALLING THE CODE

BUNDLE
Copy the code bundle for the class to the C:/Code folder.
ADDITIONAL RESOURCES
The code bundle for this book is also hosted on GitHub at:
Python.
We also have other code bundles from our rich catalog of

books and videos available at
https://github.com/PacktPublishing/. Check them out!
Chapter 1
The Importance of
Data Visualization and
Data Exploration
Learning Objectives
By the end of this chapter, you will be able to:
Explain why data visualization is

important
Calculate basic statistical values

such as median, mean, and variance
Use NumPy for data wrangling
Use pandas for data wrangling
In this chapter, you will also learn about the basic operations of
NumPy and pandas.
Introduction
Unlike machines, people are not usually equipped for
interpreting a lot of information from a random set of numbers
and messages in a given piece of data. While they may know
what the data is basically comprised of, they might need help
to understand it completely. Out of all of our logical
capabilities, we understand things best through the processing
of visual information. When data is represented visually, the
probability of understanding complex builds and numbers
increases.
Python has recently emerged as a programming language that

performs well for data analysis. Python has applications across
data science pipelines that convert data into a usable format,
analyze it, and extract useful conclusions from the data to
represent it well. It provides data visualization libraries that can
help you assemble graphical representations quickly.
In this book, you will learn how to use Python in combination

with various libraries, such as NumPy, pandas, Matplotlib,
seaborn, and geoplotlib, to create impactful data
visualizations using real-world data. Besides that, you will also
learn about the features of different types of charts and
compare their advantages and disadvantages. This will help
you choose the chart type that's suited to visualizing your data.
Once we understand the basics, we can cover more advanced

concepts, such as interactive visualizations and how Bokeh can
be used to create animated visualizations that tell a story. Upon
completion of this book, you will be able to perform data
wrangling, extract important information, and visualize your
insights descriptively.
INTRODUCTION TO DATA
VISUALIZATION
Computers and smartphones store data such as names and
numbers in a digital format. Data representation refers to the
form in which you can store, process, and transmit data.
Representations can narrate a story and convey key discoveries

to your audience. Without appropriately modeling your
information to use it to make meaningful findings, its value is
reduced. Creating representations helps to achieve a clearer,
more concise, and direct perspective of information, making it
easier for anyone to understand the data.
Information isn't really equivalent to data. Representations are

a useful apparatus to find insights hidden in data. Thus,
representations transform information into useful data.
THE IMPORTANCE OF
DATA VISUALIZATION
Visual data is very easy to understand compared to data in any
other form. Instead of just looking at data in the columns of an
Excel spreadsheet, we get a better idea of what our data
contains by using a visualization. For instance, it's easy to see a
pattern emerge from the numerical data that's given in the
following graph:
Figure 1.1: A simple example of data visualization
Visualizing data has many advantages, such as the following:
Complex data can be easily

understood
A simple visual representation of

outliers, target audiences, and
future markets can be created
Storytelling can be done using

dashboards and animations
Data can be explored through

interactive visualizations
DATA WRANGLING
To draw conclusions from visualized data, we need to handle
our data and transform it into the best possible representation.
This is where data wrangling is used. It is the discipline of
augmenting, transforming, and enriching data in a way that
allows it to be displayed and understood by machine learning
algorithms.
Look at the following data wrangling process flow diagram to

understand how accurate and actionable data can be obtained
for business analysts to work on. As you can see, the Employee
Engagement data is in its raw form initially. It gets imported as
a DataFrame and is later cleaned. The cleaned data is then
transformed into corresponding graphs, from which insights
can be modeled. Based on these insights, we can communicate
the final results. For example, employee engagement can be
measured based on raw data gathered from feedback surveys,
employee tenure, exit interviews, one-on-one meetings, and so
on. This data is cleaned and made into graphs based on
parameters such as referrals, faith in leadership, and scope of
promotions. The percentages, that is, insights generated from
the graphs, help us reach our result, which is to determine the
measure of employee engagement:
Figure 1.2: Data wrangling process to measure
employee engagement
TOOLS AND LIBRARIES

FOR VISUALIZATION
There are several approaches to creating data visualizations.
Depending on your background, you might want to use a non-
coding tool such as Tableau, which gives you the ability to get
a good feel for your data. Besides Python, which will be used
in this book, MATLAB and R are heavily used in data
analytics.
However, Python is the most popular language in the industry.

Its ease of use and the speed at which you can manipulate and
visualize data, combined with the availability of a number of
libraries, makes Python the best choice.
Note
MATLAB (https://www.mathworks.com/products/matlab.html),
R (https://www.r-project.org), and Tableau
(https://www.tableau.com) are not part of this book, so we will
only cover the highlighted tools and libraries for Python.
Overview of Statistics
Statistics is a combination of the analysis, collection,
interpretation, and representation of numerical data.
Probability is a measure of the likelihood that an event will
occur and is quantified as a number between 0 and 1.
A probability distribution is a function that provides the

probabilities for every possible event. Probability distribution
is frequently used for statistical analysis. The higher the
probability, the more likely the event. There are two types of
probability distribution, namely discrete probability
distribution and continuous probability distribution.
A discrete probability distribution shows all the values that a

random variable can take, together with their probability. The
following diagram illustrates an example of a discrete
probability distribution. If we have a 6-sided die, we can roll
each number between 1 and 6. We have six events that can
occur based on the number rolled. There is an equal probability
of rolling any of the numbers, and the individual probability of
any of the six events occurring is 1/6:
Figure 1.3: Discrete probability distribution for die
rolls
A continuous probability distribution defines the

probabilities of each possible value of a continuous random
variable. The following diagram illustrates an example of a
continuous probability distribution. This example illustrates the
distribution of the time needed to drive home. In most cases,
around 60 minutes is needed, but sometimes less time is
needed because there is no traffic, and sometimes much more
time is needed if there are traffic jams:
Figure 1.4: Continuous probability distribution for the
time taken to reach home
MEASURES OF CENTRAL
TENDENCY
Measures of central tendency are often called averages and
describe central or typical values for a probability distribution.
There are three kinds of averages that we are going to discuss
in this chapter:
Mean: The arithmetic average that

is computed by summing up all
measurements and dividing the sum
by the number of observations. The
mean is calculated as follows:
Median: This is the middle value

of the ordered dataset. If there is an
even number of observations, the
median will be the average of the
two middle values. The median is
less prone to outliers compared to
the mean, where outliers are
distinct values in data.
Mode: Our last measure of central

tendency, the mode is defined as the
most frequent value. There may be
more than one mode in cases where
multiple values are equally
frequent.
Example:
A die was rolled ten times and we got the following numbers:
4, 5, 4, 3, 4, 2, 1, 1, 2, and 1.
The mean is calculated by summing up all events and dividing

them by the number of observations:
(4+5+4+3+4+2+1+1+2+1)/10=2.7.
To calculate the median, the die rolls have to be ordered

according to their values. The ordered values are as follows: 1,
1, 1, 2, 2, 3, 4, 4, 4, 5. Since we have an even number of die
rolls, we need to take the average of the two middle values.
The average of the two middle values is (2+3)/2=2.5.
The modes are 1 and 4 since they are the two most frequent
events.
MEASURES OF
DISPERSION
Dispersion, also called variability, is the extent to which a
probability distribution is stretched or squeezed.
The different measures of dispersion are as follows:
Variance: The variance is the

expected value of the squared
deviation from the mean. It
describes how far a set of numbers
is spread out from their mean.
Variance is calculated as follows:
Standard deviation: It's the square

root of the variance.
Range: This is the difference

between the largest and smallest
values in a dataset.
Interquartile range: Also called

the midspread or middle 50%, it
is the difference between the 75th
and 25th percentiles, or between the
upper and lower quartiles.
CORRELATION
The measures we have discussed so far only considered single
variables. In contrast, correlation describes the statistical
relationship between two variables:
In positive correlation, both variables move in the same

direction
In negative correlation, the variables move in opposite

directions
In zero correlation, the variables are not related
Note
One thing you should be aware of is that correlation does not
imply causation. Correlation describes the relationship
between two or more variables, while causation describes how
one event is caused by another. For example: sleeping with
your shoes on is correlated with waking up with a headache.
This does not mean that sleeping with your shoes on causes a
headache in the morning. There might be a third, hidden
variable, for example, someone was up working late the
previous night, which caused both them falling asleep with
their shoes on and waking up with a headache.
Example:
You want to find a decent apartment to rent that is not too

expensive compared to other apartments you've found. The
other apartments you found on a website are priced as follows:
$700, $850, $1,500, and $750 per month:
The mean is
The median is
The standard deviation is
The range is
The median is a better statistical

measure in this case, since it is less
prone to outliers (the rent amount
of $1,500).
TYPES OF DATA
It is important to understand what kind of data you are dealing
with so that you can select both the right statistical measure
and the right visualization. We categorize data as
categorical/qualitative and numerical/quantitative. Categorical
data describes characteristics, for example, the color of an
object or a person's gender. We can further divide categorical
data into nominal and ordinal data. In contrast to nominal data,
ordinal data has an order.
Numerical data can be divided into discrete and continuous

data. We speak of discrete data if the data can only have certain
values, whereas continuous data can take any value (sometimes
limited to a range).
Another aspect to consider is whether the data has a temporal

domain – in other words, is it bound to time or does it change
over time? If the data is bound to a location, it might be
interesting to show the spatial relationship, so you should keep
that in mind as well:
Figure 1.5: Classification of types of data

SUMMARY STATISTICS
In real-world applications, we often encounter enormous
datasets, therefore, summary statistics are used to summarize
important aspects of data. They are necessary to communicate
large amounts of information in a compact and simple way.
We have already covered measures of central tendency and

dispersion, which are both summary statistics. It is important to
know that measures of central tendency show a center point in
a set of data values, whereas measures of dispersion show how
spread out the data values are.
The following table gives an overview of which measure of

central tendency is best suited to a particular type of data:
Figure 1.6: Best-suited measures of central tendency

for different types of data
NumPy
When handling data, we often need a way to work with
multidimensional arrays. As we discussed previously, we also
have to apply some basic mathematical and statistical
operations on that data. This is exactly where NumPy positions
itself. It provides support for large n-dimensional arrays and is
the built-in support for many high-level mathematical and
statistical operations.
Note
Before NumPy, there was a library called Numeric. However,
it's no longer used, as NumPy's signature ndarray allows for
the performant handling of large and high-dimensional
matrices.
Those ndarrays are the essence of NumPy. They are what

makes it faster than using Python's built-in lists. Other than the
built-in list datatype, ndarrays provide a stridden view of
memory (for example, int[] in Java). Since they are
homogeneously typed, meaning all the elements must be of the
same type, the stride is consistent, which results in less
memory wastage and better access times.
A stride is the number of locations between the beginnings of

two adjacent elements in an array. They are normally measured
in bytes or in units of the size of the array elements. A stride
can be larger or equal to the size of the element, but not
smaller, otherwise it would intersect the memory location of
the next element.
Note
Remember that NumPy arrays have a "defined" datatype. This
means you are not able to insert strings into an integer type
array. NumPy is mostly used with double-precision datatypes.
SAMPLE DATASET AND
Note
All exercises and activities will be developed in the Jupyter
Notebook. Please download the GitHub repository with all the
prepared templates from
Python
In this exercise, we will be loading the
normal_distribution.csv dataset and calculating the
mean of each row and each column in it:
1. Open the exercise01.ipynb

Jupyter Notebook from the
Lesson01 folder to implement
this exercise.
To do that, you need to navigate to

the path of this file. In the
command-line terminal, type
jupyter-lab.
2. You will now see a browser

window open, showing the
directory content you called the
previous command in. Click on
exercise01.ipynb. This will
open the notebook.
3. The notebook for Chapter01 should

now be open and ready for you to
modify.After a short introduction,
you should see a cell that imports
the necessary dependencies. In this
case, we will import numpy with
an alias:
# importing the
necessary dependencies
import numpy as np
4. Look for the cell that has a
comment saying, "loading the
dataset." This is the place you want
to insert the genfromtxt method
call. This method helps in loading
the data from a given text or .csv
file.
The complete line should look like

this:
# loading the dataset
dataset =
np.genfromtxt('./data/
normal_distribution.cs
v', delimiter=',')
5. If everything works as expected,

the generation should run through
without any error or output. Have a
look at the data you just imported
by simply writing the name of the
ndarray in the next cell. This is
implemented in the following code.
Simply executing a cell that returns
a value such as an ndarray GI will
use Jupyter formatting, which looks
nicer and, in most cases, displays
more information than using
print:
# looking at the
dataset
dataset
The output of the preceding code is

as follows:
Figure 1.7: The first rows of the

normal_distribution.csv file
6. To get a quick overview of our

dataset, we want to print out the
"shape" of it.
This will give us an output of the

form (rows, columns). Print out the
shape using the dataset.shape
command. We can also call the
rows as instances and the columns
as features. This means that our
dataset has 24 instances and 8
features:
# printing the shape
of our dataset
dataset.shape

as follows:
Figure 1.8: Shape of the dataset
7. Calculating the mean is the next

step once we've loaded and checked
our dataset.
The first row in a numpy array

can be accessed by simply indexing
it with zero, like this:
dataset[0]. As we mentioned
before, NumPy has some built-in
functions for calculations such as
mean. Therefore, we can simply
call np.mean() and pass in the
dataset row to get the result. The
printed output should look as
follows:
# calculating the mean
for the first row
np.mean(dataset[0])
The output of the preceding code is:

as follows:
Figure 1.9: Mean of elements in the first

row
8. We can also do the same for the

first column by using np.mean()
in combination with the column
indexing dataset[:, 0]:

for the first column
np.mean(dataset[:, 0])

as follows:
Figure 1.10: Mean of elements in the

first column
9. If we want to get the mean for

every single row, aggregated in a
list, we can make use of the axis
tools of NumPy. By simply passing
the axis parameter in the
np.mean() call, we can define
the dimension our data will be
aggregated on.
axis=0 is horizontal and axis=1

is vertical, so if we want to have the
result for each row, we need to
choose axis=1:
# mean for each row
np.mean(dataset,
axis=1)

as follows:
Figure 1.11: Mean of elements for each

row
If we want to have the result of

each column, we need to choose
axis=0.
# mean for each column
np.mean(dataset,
axis=0)

as follows:
Figure 1.12: Mean of elements for each
column
10. As the last task of this exercise, we

also want to have the mean of the
whole matrix. We could sum up all
the values we retrieved in the
previous steps, but NumPy allows
us to simply pass in the whole
dataset to do this calculation:

for the whole matrix
np.mean(dataset)

as follows:
Figure 1.13: Mean of elements for the complete

dataset
Congratulations! You are already one step closer to using

NumPy in combination with plotting libraries and creating
impactful visualizations. Since we've now covered the very
basics and went through calculating the mean in the previous
exercise, it's now up to you to solve the upcoming activity.
ACTIVITY 1: USING
NUMPY TO COMPUTE
THE MEAN, MEDIAN,
VARIANCE, AND
STANDARD DEVIATION
FOR THE GIVEN
NUMBERS
In this activity, we will use the skills we've learned to import
datasets and perform some basic calculations (mean, median,
variance, and standard deviation) to compute our tasks.
We want to consolidate our new skills and get acquainted with

NumPy:
1. Open the activity01.ipynb

this activity.
2. Now, import numpy into your

Jupyter Notebook and give it the
alias np.
3. Load the
normal_distribution.csv
dataset by using the genfromtxt
method of numpy.
4. Make sure everything works by

having a look at the ndarray.
5. Given the dataset, go in and use the
built-in numpy methods.
6. Using these methods, first calculate

the mean of the third row, the last
column, and the intersection of the
first 3 rows and first 3 columns.
7. Then, calculate the median of the

last row, the last 3 columns, and
each row.
8. Calculate the variance of each

column, the intersection of the last
2 rows, and the first 2 columns.
9. Calculate the standard deviation for

the dataset.
Note:
The solution for this activity can be

found on page 212.
Congratulations! You've completed your first activity using

NumPy. In the following activities, this knowledge will be
further consolidated.
BASIC NUMPY
OPERATIONS
In this section, we will learn basic NumPy operations such as
indexing, slicing, splitting, and iterating and implement them
in an activity.
Indexing
Indexing elements in a NumPy array, on a high level, works

the same as with built-in Python Lists. Therefore, we are able
to index elements in multi-dimensional matrices:
dataset[0] # index single element in

outermost dimension
dataset[-1] # index in reversed order in

outermost dimension
dataset[1, 1] # index single element in

two-dimensional data
dataset[-1, -1] # index in reversed

order in two-dimensional data
Slicing
Slicing has also been adapted from Python's Lists. Being able
to easily slice parts of lists into new ndarrays is very helpful
when handling large amounts of data:
dataset[1:3] # rows 1 and 2
dataset[:2, :2] # 2x2 subset of the data
dataset[-1, ::-1] # last row with

elements reversed
dataset[-5:-1, :6:2] # last 4 rows,

every other element up to index 6
Splitting
Splitting data can be helpful in many situations, from plotting

only half of your time-series data to separating test and training
data for machine learning algorithms.
There are two ways of splitting your data, horizontally and

vertically. Horizontal splitting can be done with the hsplit
method. Vertical splitting can be done with the vsplit
method:
np.hsplit(dataset, (3)) # split

horizontally in 3 equal lists
np.vsplit(dataset, (2)) # split

vertically in 2 equal lists
Iterating
Iterating the NumPy data structures, ndarrays, is also possible.

It steps over the whole list of data one after another, visiting
every single element in the ndarray once. Considering that they
can have several dimensions, indexing gets very complex.
nditer is a multi-dimensional iterator object that iterates

over a given number of arrays:
# iterating over whole dataset (each

value in each row)
for x in np.nditer(dataset):
print(x)
ndenumerate will give us exactly this index, thus returning

(0, 1) for the second value in the first row:
# iterating over whole dataset with

indices matching the position in the
dataset
for index, value in

np.ndenumerate(dataset):
print(index, value)
ITERATING
In this activity, we will use the features of NumPy to index,
slice, split, and iterate ndarrays to consolidate what we've
learned. Our client wants us to prove that our dataset is nicely
distributed around the mean value of 100:

this activity.
2. Now, import numpy into your

Jupyter Notebook and give it the
alias np.
3. Load the
dataset using NumPy. Make sure
that everything works by having a
look at the ndarray, as in the
previous activity. Follow the task
description in the notebook.
4. With the dataset loaded, use the

previously discussed indexing
feature to index the second row of
the dataset (second row), index the
last element of the dataset (last
row), index the first value of the
second row (second row, first
value), and index the last value of
the second-to-last row (using
combined access).
5. Creating sub-lists of the input
requires slicing. Slice an
intersection of four elements (2x2)
of the first two rows and first two
columns, select every second
element of the fifth row, and
reverse the entry order, selecting
the first two rows in reverse order.
6. If we need a smaller dataset, we can

also use splitting to effectively
divide our dataset. Use this concept
to split our dataset horizontally into
three equal parts and split our
dataset vertically on index 2.
7. The last task in this activity will be

iterating over the complete dataset.
Use the previously discussed
methods to iterate over the whole
dataset, with indices matching the
position in the dataset.
Note:

found on page 216.
Congratulations! We've already covered most of the basic data

wrangling methods for NumPy. In the next activity, we'll take a
look at more advanced features that will give you the tools to
get better insights into your data.
ADVANCED NUMPY
OPERATIONS
In this section, we will learn GI advanced NumPy operations
such as filtering, sorting, combining, and reshaping and
implement them in an Activity.
Filtering
Filtering is a very powerful tool that can be used to clean up

your data if you want to avoid outlier values. It's also helpful to
get some better insights into your data.
In addition to the dataset[dataset > 10] shorthand

notation, we can use the built-in NumPy extract method,
which does the same thing using a different notation, but gives
us greater control with more complex examples.
If we only want to extract the indices of the values that match

our given condition, we can use the built-in where method.
For example, np.where(dataset > 5) will return a list
of indices of the values from the initial dataset that are bigger
than 5:
dataset[dataset > 10] # values bigger

than 10
np.extract((dataset < 3), dataset) #

alternative – values smaller than 3
dataset[(dataset > 5) & (dataset < 10)]

# values bigger 5 and smaller 10
np.where(dataset > 5) # indices of

values bigger than 5 (rows and cols)
Sorting
Sorting each row of a dataset can be really useful. Using

NumPy, we are also able to sort on other dimensions, such as
columns.
In addition, argsort gives us the possibility to get a list of

indices, which would result in a sorted list:
np.sort(dataset) # values sorted on last

axis
np.sort(dataset, axis=0) # values sorted

on axis 0
np.argsort(dataset) # indices of values

in sorted list
Combining
Stacking rows and columns onto an existing dataset can be

helpful when you have two datasets of the same dimension
saved to different files.
Given two datasets, we use vstack to "stack" dataset_1 on

top of dataset_2, which will give us a combined dataset
with all the rows from dataset_1, followed by all the rows
from dataset_2.
If we use hstack, we stack our datasets "next to each other,"

meaning that the elements from the first row of dataset_1
will be followed by the elements of the first row of
dataset_2. This will be applied to each row:
np.vstack([dataset_1, dataset_2]) #
combine datasets vertically
np.hstack([dataset_1, dataset_2]) #
combine datasets horizontally
np.stack([dataset_1, dataset_2], axis=0)

# combine datasets on axis 0
Note
Combining with stacking can take some getting used to. Please
look at the examples in the NumPy documentation for further
information:https://docs.scipy.org/doc/numpy-
1.15.0/reference/generated/numpy.hstack.html.
Reshaping
Reshaping can be crucial for some algorithms. Depending on

the nature of your data, it might help you to reduce
dimensionality to make visualization easier:
dataset.reshape(-1, 2) # reshape dataset

to two columns x rows
np.reshape(dataset, (1, -1)) # reshape

dataset to one row x columns
Here, -1 is an unknown dimension that NumPy identifies

automatically. NumPy will first figure out the length of any
given array and the remaining dimensions and will thus make
sure that it satisfies the given mentioned standard.
SORTING, COMBINING,
AND RESHAPING
This last activity for NumPy provides some more complex
tasks to consolidate our learning. It will also combine most of
the previously learned methods as a recap. Perform the
following steps:

this activity.
2. NumPy will be the only necessary

dependency for this activity, so
make sure to import it.
3. Again, load the

look at the ndarray.
4. After loading the dataset, use the

previously discussed filtering
features to filter for values that are
greater than 105, filter for values
that are between 90 to 95, and get
indices of values that have a delta
(difference) of less than 1 to 100.
5. Sorting data is an important feature

when trying to display data on an
ordinal datatype column. Use
sorting to sort values for each row,
sort values for each column, get
indices of positions for each row,
and get the 3 smallest values for
each row (the remaining values are
not sorted).
6. Using the split data from Activity02

of this chapter, we can also use the
previously discussed combining
features to add the second half of
the first column back together, add
the second column to our combined
dataset, and add the third column to
our combined dataset.
7. Use the reshaping features to

reshape the dataset in a one-
dimensional list with all the values
and reshape the dataset into a
matrix with only two columns.
Note:

found on page 220.
Next, we will be learning about pandas, which will bring

several advantages when working with data that is more
complex than simple multi-dimensional numerical data. pandas
also supports different datatypes in datasets, meaning that we
can have columns that hold strings and others that have
numbers.
NumPy itself, as you've seen, has some really powerful tools.

Some of them are even more powerful when combined with
pandas.
pandas
The pandas Python library offers data structures and methods
to manipulate different types of data, such as numerical and
temporal. These operations are easy to use and highly
optimized for performance.
Data formats such as CSV, JSON, and databases can be used

for DataFrame creation. DataFrames are the internal
representation of data and are very similar to tables, but are
more powerful. Importing and reading both files and in-
memory data is abstracted into a user-friendly interface. When
it comes to handling missing data, pandas provides built-in
solutions to clean up and augment your data, meaning it fills in
missing values with reasonable values.
Integrated indexing and label-based slicing in combination

with fancy indexing (what we already saw with NumPy) makes
handling data simple. More complex techniques such as
reshaping, pivoting, and melting data, together with the
possibility of easily joining and merging data, provide
powerful tooling to handle your data correctly.
If you're working with time-series data, operations such as

date range generation, frequency conversion, and moving
window statistics can provide an advanced interface for your
wrangling.
Note
Installation instructions for pandas can be found here:
https://pandas.pydata.org/.
ADVANTAGES OF
PANDAS OVER NUMPY
The following are some of the advantages of pandas:
High level of abstraction: pandas

has a higher abstraction level than
NumPy, which gives it a simpler
interface for users to interact with.
It abstracts away some of the more
complex concepts and makes it
easier to use and understand.
Less intuition: Many methods,
such as joining, selecting, and
loading files, are usable without
much intuition and without taking
away much of the powerful nature
of pandas.
Faster processing: The internal

representation of DataFrames
allows faster processing for some
operations. Of course, this always
depends on the data and its
structure.
Easy DataFrames design:

DataFrames are designed for
operations with and on large
datasets.
DISADVANTAGES OF
PANDAS
The following are some of the disadvantages of pandas:
Less applicable: Due to its higher

abstraction, it's generally less
applicable than NumPy. Especially
when used outside of its scope,
operations quickly get very
complex and hard to do.
More disk space: Due to the
mentioned internal representation
of DataFrames and the way pandas
trades disk space for a more
performant execution, the memory
usage of complex operations can
spike.
Performance problems: Especially

when doing heavy joins, which is
not recommended, memory usage
can get critical and might lead to
performance problems.
Hidden complexity: The

comparatively simple interface has
its downsides too. Less experienced
users often tend to overuse methods
and execute them several times
instead of reusing what they've
already calculated. This hidden
complexity makes users think that
the operations themselves are
simple, which is not the case.
Note
Always try to think about how to

design your workflows instead of
excessively using operations.
SAMPLE DATASET AND
In this exercise, we will be loading the
world_population.csv dataset and calculating the mean
of some rows and columns. Our dataset holds the yearly
population density for every country. Let's use pandas to get
some really quick and easy insights:

this exercise.
2. Import the pandas libraries:
# importing the
import pandas as pd
3. After importing pandas, we can use

the read_csv method to load the
mentioned dataset. We want to use
the first column, containing the
country names, as our index. We
will use the index_col
parameter for that. The complete
line should look like this:

dataset =
pd.read_csv('./data/wo
rld_population.csv',
index_col=0)
4. As before, have a look at the data

you just imported by simply writing
the name of the dataset in the next
cell. pandas uses a data structure
called DataFrames. We want to
only print some of the rows to
avoid filling the screen. pandas
DataFrames come with two
methods, head() and tail(), to
do this for you. Both take number,
n, as a parameter, which describes
how many rows should be returned:
Note
Simply executing a cell that returns

a value such as a DataFrame will
use Jupyter formatting, which looks
nicer and, in most cases, displays
more information than using
print.
# looking at the
dataset
dataset.head()
as follows:
Figure 1.14: The first five rows of our

dataset
5. To get a quick overview of our

dataset, we want to print out its
shape.
This will give us the output in the

form (rows, columns). Print
out the shape using the
dataset.shape command. This
works exactly the same as with
NumPy ndarrays:
# printing the shape
of our dataset
dataset.shape

as follows:
Figure 1.15: The shape of the dataset
6. Calculating the mean is the next

step, once we've loaded and
checked our dataset. Indexing rows
work a little bit differently and we'll
look at this in detail in the activities
that follow. For now, we only want
to index the column with the year
1961.
pandas DataFrames have built-in

functions for calculations such as
mean. This means we can simply
call dataset.mean() to get the
result.
The printed output should look as

follows:

for 1961 column
dataset["1961"].mean()

as follows:

1961 column
7. Just to see the difference in

population density over the years,
we'll do the same with the column
for the year 2015 (the population
more than doubled in the given
time range):

for 2015 column

as follows:

2015 column
8. If we want to get the mean for

every single country (row), we can
make use of pandas axis tools. By
simply passing the axis parameter
in the dataset.mean() call, we
define the dimension our data will
be aggregated on.
axis=0 is horizontal (per column)
and axis=1 is vertical (per row),
so if we want to have the result for
each row, we need to choose
axis=1. Since we have 264 rows
in our dataset, we want to restrict
the amount of returned countries to
only 10. As we discussed before,
we can use the head(10) and
tail(10) methods with a
parameter of 10:
# mean for each
country (row)
dataset.mean(axis=1).h
ead(10)

as follows:

first 10 countries (rows)
Following is the code for tail
method:
# mean for each
feature (col)
dataset.mean(axis=0).t
ail(10)

as follows:
Figure 1.19: Mean of elements for the

last 10 years (columns)
9. The last task of this exercise is to

calculate the mean of the whole
DataFrame. Since pandas
DataFrames can have different
datatypes in each column,
aggregating this value on the whole
dataset out of the box makes no
sense. By default, axis=0 will be
used, which means that this will
give us the same result as the cell
before:

for the whole matrix
dataset.mean()

as follows:
Figure 1.20: Mean of elements for each column

Congratulations! We've now seen that the interface of pandas
has some similar methods to NumPy, which makes it really
easy to understand. We have now covered the very basics,
which will help you solve the first activity using pandas. In the
following activity, you will consolidate your basic knowledge
of pandas and use the methods you just learned to solve several
computational tasks.
ACTIVITY 4: USING
PANDAS TO COMPUTE
VARIANCE FOR THE
GIVEN NUMBERS
In this activity, we will take the previously learned skills of
importing datasets and doing some basic calculations and apply
them to solve the tasks of our first activity using pandas.
We want to consolidate our new skills and get acquainted with

pandas:
1. Open the Jupyter Notebook

activity04.ipynb from the
this activity.
2. Since we are now working with

pandas, we have to import it at the
start of our activity to be able to use
it in the notebook.
3. Load the
world_population.csv
dataset using the read_csv
method of pandas.
4. Given the dataset, go in and use

pandas' built-in methods to
calculate the mean of the third row,
the last row, and the country
Germany.
5. Then, calculate the median of the

last row, the last 3 rows, and the
first 10 countries.
6. Last, calculate the variance of the

last 5 columns.
Note:

found on page 223.
Congratulations! You've completed your first activity with

pandas, which showed you some of the similarities but also
differences when working with NumPy and pandas. In the
following activities, this knowledge will be consolidated.
You'll also be introduced to more complex features and
methods of pandas.
BASIC OPERATIONS OF
PANDAS
In this section, we will learn the basic pandas operations such
as Indexing, Slicing, and Iterating and implement it with an
Activity.
Indexing
Indexing with pandas is a bit more complex than with NumPy.

We can only access columns with the single bracket. To use the
indices of the rows to access them, we need the iloc method.
If we want to access them by index_col (which was set in
the read_csv call), we need to use the loc method:
dataset["2000"] # index the 2000 col
dataset.iloc[-1] # index the last row
dataset.loc["Germany"] # index the row

with index Germany
dataset[["2015"]].loc[["Germany"]] #
index row Germany and column 2015
Slicing
Slicing with pandas is even more powerful. We can use the

default slicing syntax we've already seen with NumPy or use
multi-selection. If we want to slice different rows or columns
by name, we can simply pass a list into the bracket:
dataset.iloc[0:10] # slice of the first

10 rows
dataset.loc[["Germany", "India"]] #
slice of rows Germany and India
# subset of Germany and India with years

1970/90
dataset.loc[["Germany", "India"]]
[["1970", "1990"]]
Iterating
Iterating DataFrames is also possible. Considering that they

can have several dimensions and dtypes, the indexing is very
high level and iterating over each row has to be done
separately:
# iterating the whole dataset
for index, row in dataset.iterrows():
print(index, row)
SERIES
A pandas Series is a one-dimensional labelled array that is
capable of holding any type of data. We can create a Series by
loading datasets from a .csv file, Excel spreadsheet, or SQL
database. There are many different ways to create them. For
example:
NumPy arrays:
# import pandas
import pandas as pd
# import numpy
import numpy as np
# creating a numpy
array
numarr =
np.array(['p','y','t',
'h','o','n'])
ser =
pd.Series(numarr)
print(ser)
pandas lists:
# import pandas
import pandas as pd
# creating a pandas
list
plist =
['p','y','t','h','o','
n']
ser = pd.Series(plist)
print(ser)
USING PANDAS
In this activity, we will use previously discussed pandas
features to index, slice, and iterate DataFrames using pandas
Series. To get some understandable insights into our dataset,
we need to be able to explicitly index, slice, and iterate our
data. For example, we can compare several countries in terms
of population density growth.
After looking at the distinct operations, we want to display the

population density of Germany, Singapore, the United States,
and India for the years 1970, 1990, and 2010:

this activity.
2. Before loading the dataset, we need

to import pandas. We'll again use
the pd alias to refer to pandas.
3. Load the
dataset using pandas. Make sure
everything works by having a look
at the DataFrames.
4. With the dataset loaded, use the

previously discussed indexing
feature to index the row for the
USA, the second to last row, the
column of the year 2000 as Series,
and the population density for India
in 2000.
5. Creating sub-lists of the dataset

requires slicing. Use this concept to
slice the countries in rows 2 to 5,
slice the countries Germany,
Singapore, the United States, and
India, and slice Germany,
Singapore, the United States, and
India with their population density
for the years 1970, 1990, and 2010.
6. The last task in this activity will be

iterating over the first three
countries in our dataset. Use the
previously discussed iteration
method to iterate over the whole
dataset and print the name, country
code, and population for the years
1970, 1990, and 2010.
Note:

found on page 228.

wrangling methods using pandas. In the next activity, we'll take
a look at more advanced features such as filtering, sorting, and
reshaping to prepare you for the next chapter.
ADVANCED PANDAS
OPERATIONS
In this section, we will learn advanced pandas operations such
as filtering, sorting, and reshaping and implement them in an
activity.
Filtering
Filtering in pandas has a higher-level interface than NumPy.

You can still use the "simple" brackets-based conditional
filtering. However, you're also able to use more complex
queries, for example, filter rows based on a regular expression:
dataset.filter(items=["1990"]) # only
column 1994
dataset[(dataset["1990"] < 10)] #

countries'population density < 10 in
1999
dataset.filter(like="8", axis=1) # years

containing an 8
dataset.filter(regex="a$", axis=0) #
countries ending with a
Sorting
Sorting each row or column based on a given row or column

will help you get better insights into your data and find the
ranking of a given dataset. With pandas, we are able to do this
pretty easily. Sorting in ascending and descending order can be
done using the parameter known as ascending. Of course, you
can do more complex sorting by providing more than one value
in the by = [ ] list. Those will then be used to sort values
for which the first value is the same:
dataset.sort_values(by=["1999"]) #
values sorted by 1999
# values sorted by 1999 descending
dataset.sort_values(by=["1994"],
ascending=False)
Reshaping
Reshaping can be crucial for easier visualization and

algorithms. However, depending on your data, this can get
really complex:
dataset.pivot(index=["1999"] *
len(dataset), columns="Country Code",
values="1999")
Note
Reshaping is a very complex topic. If you want to dive deeper
into it, this is a good resource to get started:
https://bit.ly/2SjWzaB.
SORTING, AND
RESHAPING
This last activity for pandas provides some more complex tasks
and also combines most of the methods learned previously as a
recap. After this activity, students should be able to read the
most basic pandas code and understand its logic:

this activity. Refer back to the
introduction of this chapter for
instructions on how to open a
Jupyter Notebook in JupyterLab.
2. Filtering, sorting, and reshaping are

all pandas methods. Before we can
use them, we have to import
pandas.
3. Again, load the

dataset using pandas and make sure
everything works by having a look
at the DataFrames:
Figure 1.21: Looking at the first two
columns of our dataset
4. After loading the dataset, use the

previously discussed filtering
features to get the years 1961,
2000, and 2015 as columns, and all
the countries that had a greater
population density than 500 in the
year 2000. Also, use filtering to get
a new dataset that only contains the
years 2000 and later, countries that
start with A, and countries that
contain the word "land."
5. Sorting data is an important feature

when trying to display data on an
ordinal datatype column. Use the
sorting operation to sort values in
ascending order by 1961, in
ascending order by 2015, and in
descending order by 2015.
6. If our input dataset does not suit our

needs, we have to reshape it. Use
the reshaping features to reshape
the dataset to 2015 as row and
country codes as columns.
Note:

found on page 234.
You've now completed the topic about pandas, which

concludes this chapter. We've learned about the basic tools that
help you wrangle and work with data. pandas itself is an
incredibly powerful and heavily used tool for data wrangling.
Summary
NumPy and pandas are essential tools for data wrangling. Their
user-friendly interfaces and performant implementation make
data handling easy. Even though they only provide a little
insight into our datasets, they are absolutely valuable for
wrangling, augmenting, and cleaning our datasets. Mastering
these skills will improve the quality of your visualizations.
In this chapter, we learned the basics of NumPy and pandas,

and statistics concepts. Even though the statistical concepts
covered are very basic, they are necessary to enrich our
visualizations with information that, in most cases, is not
directly provided in our datasets. This hands-on experience
will help you implement exercises and activities in the
following chapters.
In the next chapter, we will focus on the different types of
visualizations and how to decide which visualization would be
best in your case. This will give you theoretical knowledge so
that you know when to use a specific chart type and why. It
will also lay down the fundamentals of the chapters, which will
heavily focus on teaching you how to use Matplotlib and
seaborn to create the plots discussed. After we have covered
basic visualization techniques with Matplotlib and seaborn, we
will dive deeper and explore the possibilities of interactive and
animated charts, which will introduce the element of
storytelling into our visualizations.
Chapter 2
All You Need to Know
About Plots
Learning Objectives
Identify the best plot type for a

given dataset and scenario
Explain the design practices of

certain plots
Design outstanding, tangible

visualizations
In this chapter, we will learn the basics of different types of

plots.
Introduction
In this chapter, we will focus on various visualizations and
identify which visualization is best to show certain information
for a given dataset. We will describe every visualization in
detail and give practical examples, such as comparing different
stocks over time or comparing the ratings for different movies.
Starting with comparison plots, which are great for comparing
multiple variables over time, we will look at their types, such
as line charts, bar charts, and radar charts. Relation plots are
handy to show relationships among variables. We will cover
scatter plots for showing the relationship between two
variables, bubble plots for three variables, correlograms for
variable pairs, and, finally, heatmaps.
Composition plots, which are used to visualize variables that

are part of a whole, as well as pie charts, stacked bar charts,
stacked area charts, and Venn diagrams are going to be
explained. To get a deeper insight into the distribution of
variables, distribution plots are used. As a part of distribution
plots, histograms, density plots, box plots, and violin plots will
be covered. Finally, we will talk about dot maps, connection
maps, and choropleth maps, which can be categorized into geo
plots. Geo plots are useful for visualizing geospatial data.
Comparison Plots
Comparison plots include charts that are well-suited for
comparing multiple variables or variables over time. For a
comparison among items, bar charts (also called column
charts) are the best way to go. Line charts are great for
visualizing variables over time. For a certain time period (say,
less than ten time points), vertical bar charts can be used as
well. Radar charts or spider plots are great for visualizing
multiple variables for multiple groups.
LINE CHART
Line charts are used to display quantitative values over a
continuous time period and show information as a series. A
line chart is ideal for a time series, which is connected by
straight-line segments.
The value is placed on the y-axis, while the x-axis is the

timescale.
Uses:
Line charts are great for comparing
multiple variables and visualizing
trends for both single as well as
multiple variables, especially if
your dataset has many time periods
(roughly more than ten).
For smaller time periods, vertical

bar charts might be the better
choice.
The following diagram shows a trend of real-estate prices (in

million US dollars) for two decades. Line charts are well-suited
for showing data trends:
Figure 2.1: Line chart for a single variable

Example:
The following diagram is a multiple variable line chart that

compares the stock-closing prices for Google, Facebook,
Apple, Amazon, and Microsoft. A line chart is great for
comparing values and visualizing the trend of the stock. As we
can see, Amazon shows the highest growth:
Figure 2.2: Line chart showing stock trends for the

five companies
Design practices:
Avoid too many lines per chart
Adjust your scale so that the trend

is clearly visible
Note
Design practices for plots with

multiple variables. A legend should
be available to describe each
variable.
BAR CHART
The bar length encodes the value. There are two variants of bar
charts: vertical bar charts and horizontal bar charts.
Uses:
While they are both used to

compare numerical values across
categories, vertical bar charts are
sometimes used to show a single
variable over time.
The do's and the don'ts of bar charts:
Don't confuse vertical bar charts

with histograms. Bar charts
compare different variables or
categories, while histograms show
the distribution for a single
variable. Histograms will be
discussed later in this chapter.
Another common mistake is to use

bar charts to show central
tendencies among groups or
categories. Use box plots or violin
plots to show statistical measures or
distributions in these cases.
Examples:
The following diagram shows a vertical bar chart. Each bar

shows the marks out of 100 that five students obtained in a test:
Figure 2.3: Vertical bar chart using student test data
The following diagram shows a horizontal bar chart. Each bar

shows the marks out of 100 that five students obtained in a test:
Figure 2.4: Horizontal bar chart using student test
data
The following diagram compares movie ratings, giving two

different scores. The Tomatometer is the percentage of
approved critics who have given a positive review for the
movie. The Audience Score is the percentage of users who
have given a score of 3.5 or higher out of 5. As we can see,
The Martian is the only movie with both a high Tomatometer
score and Audience Score. The Hobbit: An Unexpected
Journey has a relatively high Audience Score compared to the
Tomatometer score, which might be due to a huge fan base:
Figure 2.5: Comparative bar chart
Design practices:
The axis corresponding to the

numerical variable should start at
zero. Starting with another value
might be misleading, as it makes a
small value difference look like a
big one.
Use horizontal labels, that is, as

long as the number of bars is small
and the chart doesn't look too
cluttered.
RADAR CHART
Radar charts, also known as spider or web charts, visualize
multiple variables with each variable plotted on its own axis,
resulting in a polygon. All axes are arranged radially, starting
at the center with equal distances between one another and
have the same scale.
Uses:
Radar charts are great for

comparing multiple quantitative
variables for a single group or
multiple groups.
They are also useful to show which

variables score high or low within a
dataset, making them ideal to
visualize performance
Examples:
The following diagram shows a radar chart for a single

variable. This chart displays data about a student scoring marks
in different subjects:
Figure 2.6: Radar chart for one variable (student)
The following diagram shows a radar chart for two

variables/groups. Here, the chart explains the marks that were
scored by two students in different subjects:
Figure 2.7: Radar chart for two variables (two

students)
The following diagram shows a radar chart for multiple

variables/groups. Each chart displays data about a student's
performance in different subjects:
Figure 2.8: Radar chart with faceting for multiple

variables (multiple subjects)
Design practices:
Try to display ten factors or fewer

on one radar chart to make it easier
to read.
Use faceting for multiple

variables/groups, as shown in the
preceding diagram, to maintain
clarity.
SKILL COMPARISON
You are given scores of four employees (A, B, C, and D) for
five attributes: Efficiency, Quality, Commitment, Responsible
Conduct, and Cooperation. Your task is to compare the
employees and their skills:
1. Which charts are suitable for this

task?
2. You are given the following bar and

radar charts. List the advantages
and disadvantages for both charts.
Which is the better chart for this
task in your opinion and why?
Figure 2.9: Employee skills comparison
with a bar chart
The following figure shows a radar

chart for employee skills:
Figure 2.10: Employee skills
comparison with a radar chart
3. What could be improved in the

respective visualizations?
Note:

found on page 241.
Relation Plots
Relation plots are perfectly suited to show relationships
among variables. A scatter plot visualizes the correlation
between two variables for one or multiple groups. Bubble plots
can be used to show relationships between three variables. The
additional third variable is represented by the dot size.
Heatmaps are great for revealing patterns or correlating
between two qualitative variables. A correlogram is a perfect
visualization to show the correlation among multiple variables.
SCATTER PLOT
Scatter plots show data points for two numerical variables,
displaying a variable on both axes.
Uses:
You can detect whether a

correlation (relationship) exists
between two variables.
They allow you to plot the

relationship for multiple groups or
categories using different colors.
A bubble plot, which is a variation

of the scatter plot, is an excellent
tool for visualizing the correlation
of a third variable.
Examples:
The following diagram shows a scatter plot of height and

weight of persons belonging to a single group:
Figure 2.11: Scatter plot with a single variable (one

group)
The following diagram shows the same data as in the previous

plot but differentiates between groups. In this case, we have
different groups: A, B, and C:
Figure 2.12: Scatter plot with multiple variables (three

groups)
The following diagram shows the correlation between the body

mass and the maximum longevity for various animals grouped
by their classes. There is a positive correlation between the
body mass and the maximum longevity:
Figure 2.13: Correlation between body mass and
maximum longevity for animals
Design practices:
Start both axes at zero to represent

data accurately.
Use contrasting colors for data

points and avoid using symbols for
scatter plots with multiple groups or
categories.
Variants: scatter plots with marginal histograms
In addition to the scatter plot, which visualizes the correlation

between two numerical variables, you can plot the marginal
distribution for each variable in the form of histograms to give
better insight into how each variable is distributed.
Examples:
The following diagram shows the correlation between the body

mass and the maximum longevity for animals in the Aves class.
The marginal histograms are also shown, which helps to get a
better insight into both variables:
Figure 2.14: Correlation between body mass and
maximum longevity of the Aves class with marginal
histograms
BUBBLE PLOT
A bubble plot extends a scatter plot by introducing a third
numerical variable. The value of the variable is represented by
the size of the dots. The area of the dots is proportional to the
value. A legend is used to link the size of the dot to an actual
numerical value.
Uses:
To show a correlation between
three variables.
Example:
The following diagram shows a bubble plot that highlights the

relationship between heights and age of humans:
Figure 2.15: Bubble plot showing relation between
height and age of humans
Design practices:
Design practices for the scatter plot

are also applicable to the bubble
plot.
Don't use it for very large amounts

of data, since too many bubbles
make the chart hard to read.
CORRELOGRAM
A correlogram is a combination of scatter plots and
histograms. Histograms will be discussed in detail later in this
chapter. A correlogram or correlation matrix visualizes the
relationship between each pair of numerical variables using a
scatter plot.
The diagonals of the correlation matrix represent the

distribution of each variable in the form of a histogram. You
can also plot the relationship for multiple groups or categories
using different colors. A correlogram is a great chart for
exploratory data analysis to get a feeling for your data,
especially the correlation between variable pairs.
Examples:
The following diagram shows a correlogram for height, weight,

and age of humans. The diagonal plots show a histogram for
each variable. The off-diagonal elements show scatter plots
between variable pairs:
Figure 2.16: Correlogram with single category
The following diagram shows the correlogram with data

samples separated by color into different groups:
Figure 2.17: Correlogram with multiple categories
Design practices:
Start both axes at zero to represent

data accurately.
Use contrasting colors for data

points and avoid using symbols for
scatter plots with multiple groups or
categories.
HEATMAP
A heatmap is a visualization where values contained in a
matrix are represented as colors or color saturation. Heatmaps
are great for visualizing multivariate data, where categorical
variables are placed in the rows and columns and a numerical
or categorical variable is represented as colors or color
saturation.
Uses:
Visualization of multivariate data.

Great for finding patterns in your
data.
Examples:
The following diagram shows a heatmap for the most popular

products on the Electronics category page across various e-
commerce websites:
Figure 2.18: Heatmap for popular products in the
Electronics category
Variants: annotated heatmaps
Let's see the same example the we saw previously in an

annotated heatmap:
Figure 2.19: Annotated heatmap for popular products
in the Electronics category
ACTIVITY 8: ROAD
ACCIDENTS OCCURRING
OVER TWO DECADES
You are given a diagram that gives information about the road
accidents that have occurred over the past two decades during
the months of January, April, July, and October:
1. Identify the year during which the

number of road accidents occurred
were the least.
2. For the past two decades, identify

the month for which accidents show
a marked decrease:
Figure 2.20: Total accidents over 20 years
Note:
The solution for this activity can be found on page 241.
Composition Plots
Composition plots are ideal if you think about something as a
part of a whole. For static data, you can use pie charts, stacked
bar charts, or Venn diagrams. Pie charts or donut charts help
show proportions and percentages for groups. If you need an
additional dimension, stacked bar charts are great. Venn
diagrams are the best way to visualize overlapping groups,
where each group is represented by a circle. For data that
changes over time, you can use either stacked bar charts or
stacked area charts.
PIE CHART
Pie charts illustrate numerical proportion by dividing a circle
into slices. Each arc length represents a proportion of a
category. The full circle equals to 100%. For humans, it is
easier to compare bars than arc lengths; therefore, it is
recommended to use bar charts or stacked bar charts most of
the time.
Uses:
Compare items that are part of a

whole.
Examples:
The following diagram shows a pie chart that shows different

fielding positions of the cricket ground, such as long on, long
off, third man, and fine leg:
Figure 2.21: Pie chart showing fielding positions in a
cricket ground
The following diagram shows water usage around the world:

Figure 2.22: Pie chart for global water usage
Design practices:
Arrange the slices according to

their size in increasing/decreasing
order, either in a clockwise or
anticlockwise manner.
Make sure that every slice has a

different color.
Variants: donut chart
An alternative to a pie chart is a donut chart. In contrast to pie

charts, it is easier to compare the size of slices, since the reader
focuses more on reading the length of the arcs instead of the
area. Donut charts are also more space-efficient because the
center is cut out, so it can be used to display information or
further divide groups into sub-groups.
The following figure shows a basic donut chart:

Figure 2.23: Donut chart
The following figure shows a donut chart with subgroups:
Figure 2.24: Donut chart with subgroups

Design practices:
Use the same color (that's used for

the category) for the subcategories.
Use varying brightness levels for
the different subcategories.
STACKED BAR CHART

Stacked bar charts are used to show how a category is
divided into sub-categories and the proportion of the sub-
category, in comparison to the overall category. You can either
compare total amounts across each bar or show a percentage of
each group. The latter is also referenced as a 100% stacked
bar chart and makes it easier to see relative differences
between quantities in each group.
Uses:
Compare variables that can be

divided into sub-variables.
Examples:
The following diagram shows a generic stacked bar chart with

five groups:
Figure 2.25: Stacked bar chart to show sales of
laptops and mobiles
The following diagram shows a 100% stacked bar chart with

the same data that was used in the preceding diagram:
Figure 2.26: 100% stacked bar chart to show sales of
laptops, PCs, and mobiles
The following diagram illustrates the daily total sales of a

restaurant over several days. The daily total sales of non-
smokers are stacked on top of the daily total sales of smokers:
Figure 2.27: Daily total sales of restaurant
categorized by smokers and non-smokers
Design practices:
Use contrasting colors for stacked

bars.
Ensure that the bars are adequately

spaced to eliminate visual clutter.
The ideal space guideline between
each bar is half the width of a bar.
Categorize data alphabetically,

sequentially, or by value to order it
uniformly and make things easier
for your audience.
STACKED AREA CHART

Stacked area charts show trends for part-of-a-whole relations.
The values of several groups are illustrated on top of one
another. It helps to analyze both individual and overall trend
information.
Uses:
Show trends for time series that are

part of a whole.
Examples:
The following diagram shows a stacked area chart with the net
profits of companies like Google, Facebook, Twitter, and
Snapchat over a decade:
Figure 2.28: Stacked area chart to show net profits of

four companies
Design practices:
Using transparent colors might
improve information visibility.
ACTIVITY 9:
SMARTPHONE SALES
UNITS
You want to compare smartphone sales units for the five
biggest smartphone manufacturers over time and see whether
there is any trend:
1. Looking at the following line chart,

analyze the sales of each
manufacturer and identify the one
whose performance is exceptional
in the fourth quarter when
compared to the third quarter.
2. Analyze the performance of all

manufacturers, and make a
prediction about two companies
whose sales unit will show a
downward trend and an upward
trend:
Figure 2.29: Line chart of smartphone sales units
Note:
VENN DIAGRAM
Venn diagrams, also known as set diagrams, show all
possible logical relations between a finite collections of
different sets. Each set is represented by a circle. The circle
size illustrates the importance of a group. The size of an
overlap represents the intersection between multiple groups.
Uses:
To show overlaps for different sets
Example:
Visualizing the intersection of the

following diagram shows a Venn
diagram for students in two groups
taking the same class in a semester:
Figure 2.30: Venn diagram to show students taking
the same class
Design practices:
It is not recommended to use Venn

diagrams if you have more than
three groups. It would become
difficult to understand.
Distribution Plots
Distribution plots give a deep insight into how your data is
distributed. For a single variable, a histogram is well-suited.
For multiple variables, you can either use a box plot or a violin
plot. The violin plot visualizes the densities of your variables,
whereas the box plot just visualizes the median, the
interquartile range, and the range for each variable.
HISTOGRAM
A histogram visualizes the distribution of a single numerical
variable. Each bar represents the frequency for a certain
interval. Histograms help get an estimate of statistical
measures. You see where values are concentrated and you can
easily detect outliers. You can either plot a histogram with
absolute frequency values or alternatively normalize your
histogram. If you want to compare distributions of multiple
variables, you can use different colors for the bars.
Uses:
Get insights into the underlying

distribution for a dataset
Example:
The following diagram shows the distribution of the

Intelligence Quotient (IQ) for a test group. The solid line
indicates the mean and the dashed lines indicate the standard
deviation:
Figure 2.31: Distribution of Intelligence Quotient (IQ)

for a test group of a hundred adults
Design practices:
Try different numbers of bins, since
the shape of the histogram can vary
significantly.
DENSITY PLOT
A density plot shows the distribution of a numerical variable.
It is a variation of a histogram that uses kernel smoothing,
allowing for smoother distributions. An advantage they have
over histograms is that density plots are better at determining
the distribution shape, since the distribution shape for
histograms heavily depends on the number of bins (data
intervals).
Uses:
You can compare the distribution of

several variables by plotting the
density on the same axis and using
different colors.
Example:
The following diagram shows a basic density plot:

Figure 2.32: Density plot
The following diagram shows a basic multi-density plot:

Figure 2.33: Multi-density plot
Design practices:
Use contrasting colors for plotting

the density of multiple variables.
BOX PLOT
The box plot shows multiple statistical measurements. The box
extends from the lower to the upper quartile values of the data,
thus allowing us to visualize the interquartile range. The
horizontal line within the box denotes the median. The
whiskers extending from the box show the range of the data. It
is also an option to show data outliers, usually as circles or
diamonds, past the end of the whiskers.
Uses:
If you want to compare statistical

measures for multiple variables or
groups, you can simply plot
multiple boxes next to one another.
Examples:
The following diagram shows a basic box plot:

Figure 2.34: Box plot showing single variable
The following diagram shows a basic box plot for multiple

variables:
Figure 2.35: Box plot for multiple variables
VIOLIN PLOT
Violin plots are a combination of box plots and density plots.
Both the statistical measures and the distribution are
visualized. The thick black bar in the center represents the
interquartile range, the thin black line shows the 95%
confidence interval, and the white dot shows the median. On
both sides of the center-line, the density is visualized.
Uses:
If you want to compare statistical

measures for multiple variables or
groups, you can simply plot
multiple violins next to one another.
Examples:
The following diagram shows a violin plot for a single variable

and shows how students have performed in Math:
Figure 2.36: Violin plot for a single variable (math)
The following diagram shows a violin plot for two variables

and shows the performance of students in English and Math:
Figure 2.37: Violin plot for multiple variables (English
and math)
The following diagram shows a violin plot for a single variable

divided into three groups and shows the performance of three
divisions of students in English:
Figure 2.38: Violin plot with multiple categories (three
groups of students)
Design practices:
Scale the axes accordingly so that

the distribution is clearly visible
and not flat.

OF TRAINS DURING
DIFFERENT TIME
INTERVALS
You are provided with a histogram that states the total number
of trains arriving at different time intervals:
1. By looking at the following graph,

can you identify the interval during
which the most number of trains
arrive?
2. How would the histogram change if

the number of trains arriving
between 4 and 6 pm were to be
increased by 50?
Figure 2.39: Frequency of trains during different time
intervals
Note:
Geo Plots
Geological plots are a great way to visualize geospatial data.
Choropleth maps can be used to compare quantitative values
for different countries, states, and so on. If you want to show
connections between different locations, connection maps are
the way to go.
DOT MAP
In a dot map, each dot represents a certain number of
observations. Each dot has the same size and value (the
number of observations each dot represents). The dots are not
meant to be counted—they are only intended to give an
impression of magnitude. The size and value are important
factors for the effectiveness and impression of the
visualization. You can use different colors or symbols for the
dots to show multiple categories or groups.
Uses:
For the visualization of geospatial

data
Example:
The following diagram shows a dot map where each dot

represents a certain amount of bus stops throughout the world:
Figure 2.40: Dot map showing bus stops worldwide
Design practices:
Do not show too many locations.

You should still be able to see the
map to get a feel for the actual
location.
Choose a dot size and value so that

in dense areas, the dots start to
blend. The dot map should give a
good impression of the underlying
spatial distribution.
CHOROPLETH MAP
In a choropleth map, each tile is colored to encode a variable.
A tile represents a geographic region for, for example, counties
and countries. Choropleth maps provide a good way to show
how a variable varies across a geographic area. One thing to
keep in mind for choropleth maps is that the human eye
naturally gives more prominence to larger areas, so you might
want to normalize your data by dividing the map area-wise.
Uses:
For the visualization of geospatial

data grouped into geological
regions, for example, states, or
countries
Example:
The following diagram shows a choropleth map of a weather

forecast in the USA:
Figure 2.41: Choropleth map showing weather
forecast of USA
Design practices:
Use darker colors for higher values,

as they are perceived as being
higher in magnitude.
Limit the color gradation, since the

human eye is limited to how many
colors it can easily distinguish
between. Seven color gradations
should be enough.
CONNECTION MAP
In a connection map, each line represents a certain number of
connections between two locations. The link between the
locations can be drawn with a straight or rounded line
representing the shortest distance between them.
Each line has the same thickness and value (number of

connections each line represents). The lines are not meant to be
counted; they are only intended to give an impression of
magnitude. The size and value of a connection line are
important factors for the effectiveness and impression of the
visualization.
You can use different colors for the lines to show multiple
categories or groups, or you can use a colormap to encode the
length of the connection.
Uses:
For the visualization of connections
Examples:
The following diagram shows a connection map of flight

connections around the world:
Figure 2.42: Connection map showing Flight
connections around the world
Design practices:
Do not show too many connections.

You should still see the map to get a
feel of the actual locations of the
start and end points.
Choose a line thickness and value

so that the lines start to blend in
dense areas. The connection map
should give a good impression of
the underlying spatial distribution.
What Makes a Good

Visualization?
There are multiple aspects to what makes a good visualization:
Most importantly, a visualization

should be self-explanatory and
visually appealing. To make it self-
explanatory, use a legend,
descriptive labels for your x-axis
and y-axis, and titles.
A visualization should tell a story

and be designed for your audience.
Before creating your visualization,
think about your target audience—
create simple visualizations for a
non-specialist audience and more
technical detailed visualizations for
a specialist audience. Think about a
story to tell with your visualization
so that your visualization leaves an
impression on the audience.
Common design practices:
Colors are more perceptible than

symbols.
To show additional variables on a

2D plot, use color, shape, and size.
Keep it simple and don't overload

the visualization with too much
information.

THE IDEAL
VISUALIZATION
The following visualizations are not ideal as they do not
represent data well. Answer the following questions for each
visualization:
1. What are the bad aspects of these

visualizations?
2. How could we improve the

visualizations? Sketch the right
visualization for both scenarios.
The first visualization is supposed to illustrate the top 30

YouTubers according to the number of subscribers:
Figure 2.43: Pie chart showing Top 30 YouTubers
The second visualization is supposed to illustrate the number

of people playing a certain game in a casino over two days:
Figure 2.44: Line chart displaying casino data for two
days
Note:
Summary
In this chapter, the most important visualizations were
discussed. The visualizations were categorized into
comparison, relation, composition, distribution, and geological
plots. For each plot, a description, practical examples, and
design practices were given. Comparison plots, such as line
charts, bar charts, and radar charts, are well-suited for
comparing multiple variables or variables over time. Relation
plots are perfectly suited to show relationships between
variables. Scatter plots, bubble plots, which are an extension of
scatter plots, correlograms, and heatmaps were considered.
Composition plots are ideal if you think about something as a
part of a whole. We first covered pie charts and continued with
stacked bar charts, stacked area charts, and Venn diagrams. For
distribution plots that give a deep insight into how your data is
distributed, histograms, density plots, box plots, and violin
plots were considered. Regarding geospatial data, we discussed
dot maps, connection maps, and choropleth maps. Finally,
some remarks were given on what makes a good visualization.
In the next chapter, we will dive into Matplotlib and create our
own visualizations. We will cover all the plots that we have
discussed in this chapter.
Chapter 3
A Deep Dive into
Matplotlib
Learning Objectives
Describe the fundamentals of

Matplotlib
Create visualizations using the

built-in plots that are provided by
Matplotlib
Customize your visualization plots
Write mathematical expressions

using TeX
In this chapter, we will learn how to customize you

visualization using Matplotlib.
Introduction
Matplotlib is probably the most popular plotting library for
Python. It is used for data science and machine learning
visualizations all around the world. John Hunter began
developing Matplotlib in 2003. It aimed to emulate the
commands of the MATLAB software, which was the scientific
standard back then. Several features such as the global style of
MATLAB were introduced into Matplotlib to make the
transition to Matplotlib easier for MATLAB users.
Before we start working with Matplotlib to create our first

visualizations, we will understand and try to grasp the concepts
behind the plots.
Overview of Plots in
Matplotlib
Plots in Matplotlib have a hierarchical structure that nests
Python objects to create a tree-like structure. Each plot is
encapsulated in a Figure object. This Figure is the top-
level container of the visualization. It can have multiple axes,
which are basically individual plots inside this top-level
container.
Going a level deeper, we again find Python objects that control

axes, tick marks, legend, title, textboxes, the grid, and many
other objects. All of these objects can be customized.
The two main components of a plot are as follows:
Figure
The Figure is an outermost

container and is used as a canvas to
draw on. It allows you to draw
multiple plots within it. It not only
holds the Axes object but also has
the capability to configure the Title.
Axes
The Axes is an actual plot, or
subplot, depending on whether you
want to plot single or multiple
visualizations. Its sub-objects
include the x and y axis, spines, and
legends.
Observing this design on a higher level, we can see that this

hierarchical structure allows us to create a complex and
customizable visualization.
When looking at the "anatomy" of a Figure, which is shown in

the following diagram, we get an idea about the complexity of
an insightful visualization. Matplotlib gives us the ability not
only to simply display data, but also design the whole Figure
around it, by adjusting the Grid, x and y ticks, tick labels, and
the Legend. This implies that we can modify every single bit
of a plot, starting from the Title and Legend, right down to
even the major and minor ticks on the spines to make it more
expressive:
Figure 3.1: Anatomy of a Matplotlib Figure
Taking a deeper look into the anatomy of a Figure object, we

can observe the following components:
Spines: Lines connecting the axis

tick marks
Title: Text label of the whole

Figure object
Legend: They describe the content

of the plot
Grid: Vertical and horizontal lines
used as an extension of the tick
marks
X/Y axis label: Text label for the

X/Y axis below the spines
Minor tick: Small value indicators

between the major tick marks
Minor tick label: Text label that

will be displayed at the minor ticks
Major tick: Major value indicators

on the spines
Major tick label: Text label that

will be displayed at the major ticks
Line: Plotting type that connects

data points with a line
Markers: Plotting type that plots

every data point with a defined
marker
In this book, we will focus on Matplotlib's submodule, pyplot,

which provides MATLAB-like plotting.
Pyplot Basics
pyplot contains a simpler interface for creating visualizations,
which allows the users to plot the data without explicitly
configuring the Figure and Axes themselves. They are
implicitly and automatically configured to achieve the desired
output. It is handy to use the alias plt to reference the
imported submodule, as follows:
import matplotlib.pyplot as plt
The following sections describe some of the common

operations that are performed when using pyplot.
CREATING FIGURES
We use plt.figure() to create a new Figure. This
function returns a Figure instance, but it is also passed to the
backend. Every Figure-related command that follows is applied
to the current Figure and does not need to know the Figure
instance.
By default, the Figure has a width of 6.4 inches and a height of

4.8 inches with a dpi of 100. To change the default values of
the Figure, we can use the parameters figsize and dpi.
The following code snippet shows how we can manipulate a

Figure:
plt.figure(figsize=(10, 5)) #To change

the width and the height
plt.figure(dpi=300) #To change the dpi
CLOSING FIGURES
Figures that are not used anymore should be closed by
explicitly calling plt.close(), which also cleans up
memory efficiently.
If nothing is specified, the current Figure will be closed. To

close a specific Figure, you can either provide a reference to a
Figure instance or provide the Figure number. To find the
number of a figure object, we can make use of the number
attribute, like so:
plt.gcf().number
By using plt.close('all'), all Figures will be closed.

The following example shows how a Figure can be created and
closed:
plt.figure(num=10) #Create Figure with

Figure number 10
plt.close(10) #Close Figure with Figure

number 10
FORMAT STRINGS
Before we actually plot something, let's quickly discuss format
strings. They are a neat way to specify colors, marker types,
and line styles. A format string is specified as "[color]
[marker][line]", where each item is optional. If the
color is the only argument of the format string, you can use
any matplotlib.colors. Matplotlib recognizes the
following formats, among others:
RGB or RGBA float tuples (for

example, (0.2, 0.4, 0.3) or (0.2, 0.4,
0.3, 0.5))
RGB or RGBA hex strings (for

example, '#0F0F0F' or
'#0F0F0F0F')
The following diagram is an example of how a color can be

represented in one particular format:
Figure 3.2 Color specified in string format
All the available marker options are illustrated in the following

diagram:
Figure 3.3: Markers in format strings
All the available line styles are illustrated in the following

diagram:
Figure 3.4: Line styles
PLOTTING
With plt.plot([x], y, [fmt]), you can plot data
points as lines and/or markers. The function returns a list of
Line2D objects representing the plotted data. By default, if you
do not provide a format string, the data points will be
connected with straight, solid lines. plt.plot([0, 1, 2,
3], [2, 4, 6, 8]) produces a plot, as shown in the
following diagram. Since x is optional and default values are
[0, …, N-1], plt.plot([2, 4, 6, 8]) results in
the same plot:
Figure 3.5: Plotting data points as a line
If you want to plot markers instead of lines, you can just

specify a format string with any marker type. For example,
plt.plot([0, 1, 2, 3], [2, 4, 6, 8], 'o')
displays data points as circles, as shown in the following
diagram:
Figure 3.6: Plotting data points with markers (circles)
To plot multiple data pairs, the syntax plt.plot([x], y,

[fmt], [x], y2, [fmt2], …) can be used.
plt.plot([2, 4, 6, 8], 'o', [1, 5, 9, 13],
's') results in the following diagram. Similarly, you can use
plt.plot multiple times, since we are working on the same
Figure and Axes:
Figure 3.7: Plotting data points with multiple markers
Any Line2D properties can be used instead of format strings to

further customize the plot. For example, the following code
snippet shows how we can additionally specify the
linewidth and the markersize:
plt.plot([2, 4, 6, 8], color='blue',

marker='o', linestyle='dashed',
linewidth=2, markersize=12)
PLOTTING USING
PANDAS DATAFRAMES
It is pretty straightforward to use pandas.DataFrame as a
data source. Instead of providing x and y values, you can
provide the pandas.DataFrame in the data parameter and
give keys for x and y, as follows:
plt.plot('x_key', 'y_key', data=df)
DISPLAYING FIGURES
plt.show() is used to display a Figure or multiple Figures.
To display Figures within a Jupyter Notebook, simply set the
%matplotlib inline command in the beginning of the
code.
SAVING FIGURES
plt.savefig(fname) saves the current Figure. There are
some useful optional parameters you can specify, such as dpi,
format, or transparent. The following code snippet
gives an example of how you can save a Figure:
plt.figure()
plt.plot([1, 2, 4, 5], [1, 3, 4, 3], '-

o')
plt.savefig('lineplot.png', dpi=300,
bbox_inches='tight')
#bbox_inches='tight' removes the outer
white margins
Note
All exercises and activities will be developed in the Jupyter
Notebook. Please download the GitHub repository with all the
prepared templates from
Python.
SIMPLE VISUALIZATION
In this exercise, we will create our first simple plot using
Matplotlib:

exercise03.ipynb from the
this exercise.
Navigate to the path of this file and

type in the following at the
command line: jupyter-lab.
2. Import the necessary modules and

enable plotting within a Jupyter
Notebook:
import numpy as np
import
matplotlib.pyplot as
plt
%matplotlib inline
3. Explicitly create a Figure and set

the dpi to 200:
plt.figure(dpi=200)
4. Plot the following data pairs (x,

y) as circles, which are connected
via line segments: (1, 1), (2,
3), (4, 4), (5, 3) . Then,
visualize the plot:
plt.plot([1, 2, 4, 5],
[1, 3, 4, 3], '-o')
plt.show()
Your output should look similar to

this:
Figure 3.8: A simple visualization that

was created with the help of given data
pairs and is connected via line
segments
5. Save the plot using the

plt.savefig() method. Here,
we can either provide a filename
within the method or specify the
full path:
plt.savefig(exercise03
.png);
Basic Text and

Legend Functions
All of the functions we have discussed in this topic, except for
the legend, create and return a
matplotlib.text.Text() instance. We are mentioning
it here so that you know that all of the discussed properties can
be used for the other functions as well. All text functions are
illustrated in Figure 3.9.
LABELS
Matplotlib provides a few label functions that we can use for
setting labels to the x and y axes. The plt.xlabel() and
plt.ylabel() functions are used to set the label for the
current axes. The set_xlabel() and set_ylabel()
functions are used to set the label for specified axes.
Example:
ax.set_xlabel('X Label')
ax.set_ylabel('Y Label')
TITLES
A title describes a particular chart/graph. The titles are
placed above the axes in the center, left edge, or right edge.
There are two options for titles – you can either set the Figure
title or the title of an Axes. The suptitle() function sets
the title for current and specified Figure. The title()
function helps in setting the title for the current and specified
axes.
Example:
fig = plt.figure()
fig.suptitle('Suptitle', fontsize=10,
fontweight='bold')
This creates a bold figure title with a text suptitle and a font
size of 10.
TEXT
There are two options for text – you can either add text to a
Figure or text to an Axes. The figtext(x, y, text) and
text(x, y, text) functions add a text at location x, or y
for a figure.
Example:
ax.text(4, 6, 'Text in Data Coords',

bbox={'facecolor': 'yellow',
'alpha':0.5, 'pad':10})
This creates a yellow textbox with the text "Text in Data

Coords".
ANNOTATIONS
Compared to text that is placed at an arbitrary position on the
Axes, annotations are used to annotate some features of the
plot. In annotation, there are two locations to consider: the
annotated location xy and the location of the annotation, text
xytext. It is useful to specify the parameter arrowprops,
which results in an arrow pointing to the annotated location.
Example:
ax.annotate('Example of Annotate', xy=

(4,2), xytext=(8,4),
arrowprops=dict(facecolor='green',
shrink=0.05))
This creates a green arrow pointing to the data coordinates (4,

2) with the text "Example of Annotate" at data coordinates (8,
4):
Figure 3.9: Implementation of Text commands

LEGENDS
For adding a legend to your Axes, we have to specify the
label parameter at the time of artist creation. Calling
plt.legend() for the current Axes or Axes.legend()
for a specific Axes will add the legend. The loc parameter
specifies the location of the legend.
Example:
plt.plot([1, 2, 3], label='Label 1')
plt.plot([2, 4, 3], label='Label 2')
plt.legend()
This example is illustrated in the following diagram:
Figure 3.10: Legend example

STOCK TRENDS BY
USING A LINE PLOT
In this activity, we will create a line plot to show stock trends.
Let's look at the following scenario: you are interested in
investing in stocks. You downloaded the stock prices for the
"big five": Amazon, Google, Apple, Facebook, and Microsoft:
1. Use pandas to read the data located

in the subfolder data.
2. Use Matplotlib to create a line chart

visualizing the closing prices for
the past five years (whole data
sequence) for all five companies.
Add labels, titles, and a legend to
make the visualization self-
explanatory. Use plt.grid() to
add a grid to your plot.
3. After executing the preceding steps,

the expected output should be as
follows:
Figure 3.11: Visualization of stock trends of five
companies
Note:
Basic Plots
In this section, we are going to go through the different types
of basic plots.
BAR CHART
plt.bar(x, height, [width]) creates a vertical bar
plot. For horizontal bars, use the plt.barh() function.
Important parameters:
x: Specifies the x coordinates of the

bars
height: Specifies the height of

the bars
width (optional): Specifies the

width of all bars; the default is 0.8
Example:
plt.bar(['A', 'B', 'C', 'D'], [20, 25,

40, 10])
The preceding code creates a bar plot, as shown in the

following diagram:
Figure 3.12: A simple bar chart
If you want to have subcategories, you have to use the

plt.bar() function multiple times with shifted x-
coordinates. This is done in the following example and
illustrated in the diagram that follows. The arange()
function is a method in the NumPy package that returns evenly
spaced values within a given interval. The gca() function
helps in getting the instance of current axes on any current
figure. The set_xticklabels() function is used to set the
x-tick labels with the list of given string labels.
Example:
labels = ['A', 'B', 'C', 'D']
x = np.arange(len(labels))
width = 0.4
plt.bar(x – width / 2, [20, 25, 40, 10],

width=width)
plt.bar(x – width / 2, [30, 15, 30, 20],

width=width)
# Ticks and tick labels must be set

manually
plt.ticks(x)
ax = plt.gca()
ax.set_xticklabels(labels)
The preceding code creates a bar chart with subcategories, as

shown in the following diagram:
Figure 3.13: Bar chart with subcategories

BAR PLOT FOR MOVIE
COMPARISON
In this activity, we will use a bar plot to compare movie scores.
You are given five movies with scores from Rotten Tomatoes.
The Tomatometer is the percentage of approved Tomatometer
critics who have given a positive review for the movie. The
Audience Score is the percentage of users who have given a
score of 3.5 or higher out of 5. Compare these two scores
among the five movies:

2. Use Matplotlib to create a visually

appealing bar plot comparing the
two scores for all five movies.
3. Use the movie titles as labels for

the x-axis. Use percentages in an
interval of 20 for the y-axis and
minor ticks in an interval of 5. Add
a legend and a suitable title to the
plot.

follows:
Figure 3.14: Bar plot comparing scores of five
movies
Note:
PIE CHART
The plt.pie(x, [explode], [labels],
[autopct]) function creates a pie chart.
x: Specifies the slice sizes.

explode (optional): Specifies the
fraction of the radius offset for each
slice. The explode-array must have
the same length as the x-array.
labels (optional): Specifies the

labels for each slice.
autopct (optional): Shows

percentages inside the slices
according to the specified format
string. Example: '%1.1f%%'.
Example:
plt.pie([0.4, 0.3, 0.2, 0.1], explode=

(0.1, 0, 0, 0), labels=['A', 'B', 'C',
'D'])
The result of the preceding code is visualized in following

diagram:
Figure 3.15: Basic pie chart
PIE CHART FOR WATER
USAGE
In this exercise, we will use a pie chart to visualize water
usage:

this exercise.


Notebook:
# Import statements
import pandas as pd
import
plt
%matplotlib inline

in the subfolder data:
# Load dataset
data =
pd.read_csv('./data/wa
ter_usage.csv')
4. Use a pie chart to visualize the

water usage. Highlight one usage of
your choice using the explode
parameter. Show the percentages
for each slice and add a title:
# Create figure
plt.figure(figsize=(8,
8), dpi=300)
# Create pie plot
plt.pie('Percentage',
explode=(0, 0, 0.1, 0,
0, 0), labels='Usage',
data=data,
autopct='%.0f%%')
# Add title
plt.title('Water
usage')
# Show plot
plt.show()
The following Figure shows output of the preceding code:

Figure 3.16: Pie chart for water usage
STACKED BAR CHART

A stacked bar chart uses the same plt.bar function as bar
charts. For each stacked bar, the plt.bar function must be
called and the bottom parameter must be specified starting
with the second stacked bar. This will become clear with the
following example:
plt.bar(x, bars1)
plt.bar(x, bars2, bottom=bars1)
plt.bar(x, bars3, bottom=np.add(bars1,

bars2))
The result of the preceding code is visualized in the following

diagram:
Figure 3.17: Stacked bar chart

STACKED BAR PLOT TO
PERFORMANCE
In this activity, we will use a stacked bar plot to visualize the
performance of a restaurant. Let's look at the following
scenario: you are the owner of a restaurant and, due to a new
law, you have to introduce a No Smoking Day. To make as few
losses as possible, you want to visualize how many sales are
made every day, categorized by smokers and non-smokers:
1. Use the given dataset and create a

matrix where the elements contain
the sum of the total bills for each
day and smokers/non-smokers.
2. Create a stacked bar plot, stacking

the summed total bills separated by
smoker and non-smoker for each
day. Add a legend, labels, and a
title.

follows:
Figure 3.18: Stacked bar chart showing performance
of restaurant on different days
Note:
STACKED AREA CHART

plt.stackplot(x, y) creates a stacked area plot.
x: Specifies the x-values of the data

series.
y: Specifies the y-values of the data

series. For multiple series, either as
a 2d array, or any number of 1D
arrays, call the following function:
plt.stackplot(x, y1, y2,
y3, …).
labels (Optional): Specifies the

labels as a list or tuple for each data
series.
Example:
plt.stackplot([1, 2, 3, 4], [2, 4, 5,

8], [1, 5, 4, 2])
The result of the preceding code is shown in the following

diagram:
Figure 3.19: Stacked area chart

SMARTPHONE SALES
AREA CHART
In this activity, we will compare smartphone sales units using a
stacked area chart. Let's look at the following scenario: you
want to invest in one of the biggest five smartphone
manufacturers. Looking at the quarterly sales units as part of a
whole may be a good indicator for which company to invest in:

2. Create a visually appealing stacked

area chart. Add a legend, labels,
and a title.

follows:
Figure 3.20: Stacked area chart comparing sales

units of different smartphone manufacturers
Note:
HISTOGRAM
plt.hist(x) creates a histogram.
x: Specifies the input values
bins: (optional): Either specifies

the number of bins as an integer or
specifies the bin edges as a list
range: (optional): Specifies the

lower and upper range of the bins
as a tuple
density: (optional): If true, the

histogram represents a probability
density
Example:
plt.hist(x, bins=30, density=True)

diagram:
Figure 3.21: Histogram
plt.hist2d(x, y) creates a 2D histogram. An example

of a 2D historgram is shown in the following diagram:
Figure 3.22: 2D histogram with color bar
BOX PLOT
plt.boxplot(x) creates a box plot.
x: Specifies the input data. It

specifies either a 1D array for a
single box or a sequence of arrays
for multiple boxes.
notch: Optional: If true, notches

will be added to the plot to indicate
the confidence interval around the
median.
labels: Optional: Specifies the
labels as a sequence.
showfliers: Optional: By
default, it is true, and outliers are
plotted beyond the caps.
showmeans: Optional: If true,

arithmetic means are shown.
Example:
plt.boxplot([x1, x2], labels=['A', 'B'])

diagram:
Figure 3.23: Box plot

HISTOGRAM AND A BOX
In this activity, we will visualize the intelligence quotient (IQ)
using a histogram and box plots.
Note
plt.axvline(x, [color=…], [linestyle=…])
draws a vertical line at position x.
1. Plot a histogram with 10 bins for

the given IQ scores. IQ scores are
normally distributed with a mean of
100 and a standard deviation of 15.
Visualize the mean as a vertical
solid red line, and the standard
deviation using dashed vertical
lines. Add labels and a title.
2. Create a box plot to visualize the

same IQ scores. Add labels and a
title.
3. Create a box plot for each of the IQ

scores of the different test groups.
Add labels and a title.
4. The expected output for step 1 is as
follows:
Figure 3.24: Histogram for IQ test

follows:
Figure 3.25: Box plot for IQ scores

follows:
Figure 3.26: Box plot for IQ scores of different test
groups
Note:
SCATTER PLOT
plt.scatter(x, y) creates a scatter plot of y versus x
with optionally varying marker size and/or color.
x, y: Specifies the data positions.

s: Optional: Specifies the marker
size in points squared.
c: Optional: Specifies the marker

color. If a sequence of numbers is
specified, the numbers will be
mapped to colors of the color map.
Example:
plt.scatter(x, y)

diagram:
Figure 3.27: Scatter plot

SCATTER PLOT TO
BETWEEN VARIOUS
ANIMALS
In this activity, we will use a scatter plot to show correlation
within a dataset. Let's look at the following scenario: you are
given a dataset containing information about various animals.
Visualize the correlation between the various animal attributes:
Note
Axes.set_xscale('log') and
Axes.set_yscale('log') change the scale of the x-axis
and y-axis to a logarithmic scale, respectively.
1. The given dataset is not complete.

Filter the data so that you end up
with samples containing a body
mass and a maximum longevity.
Sort the data according to the
animal class.
2. Create a scatter plot visualizing the

correlation between the body mass
and the maximum longevity. Use
different colors for grouping data
samples according to their class.
Add a legend, labels, and a title.
Use a log scale for both the x-axis
and y-axis.

follows:
Figure 3.28: Scatter plot on animal statistics
Note:
BUBBLE PLOT
The plt.scatter function is used to create a bubble plot.
To visualize a third or a fourth variable, the parameters s
(scale) and c (color) can be used.
Example:
plt.scatter(x, y, s=z*500, c=c,

alpha=0.5)
plt.colorbar()

diagram:
Figure 3.29: Bubble plot with color bar
Layouts
There are multiple ways to define a visualization layout in
Matplotlib. We will start with subplots and how to use the
tight layout to create visually appealing plots and then cover
GridSpec, which offers a more flexible way to create multi-
plots.
SUBPLOTS
It is often useful to display several plots next to each other.
Matplotlib offers the concept of subplots, which are multiple
Axes within a Figure. These plots can be grids of plots, nested
plots, and so forth.
Explore the following options to create subplots:
plt.subplots(nrows,
ncols) creates a Figure and a set
of subplots.
plt.subplot(nrows,
ncols, index) or equivalently
plt.subplot(pos) adds a
subplot to the current Figure. The
index starts at 1.
plt.subplot(2, 2, 1) is
equivalent to
plt.subplot(221).
Figure.subplots(nrows,
ncols) adds a set of subplots to
the specified Figure.
Figure.add_subplot(nrows
, ncols, index) or
equivalently
Figure.add_subplot(pos)
adds a subplot to the specified
Figure.
For sharing the x or y axis, the parameters sharex and

sharey must be set, respectively. The axis will have the same
limits, ticks, and scale.
plt.subplot and Figure.add_subplot has the option

to set a projection. For a polar projection, either set the
projection='polar' parameter or set the parameter
polar=True parameter.
Example 1:
fig, axes = plt.subplots(2, 2)
axes = axes.ravel()
for i, ax in enumerate(axes):
ax.plot(series[i])
for i in range(4):
plt.subplot(2, 2, i+1)
plt.plot(series[i])
Both examples yield the same result, as shown in the following

diagram:
Figure 3.30: Subplots
Example 2:
fig, axes = plt.subplots(2, 2,

sharex=True, sharey=True)
axes = axes.ravel()
ax.plot(series[i])
Setting sharex and sharey to true results in the following

diagram. This allows for a better comparison:
Figure 3.31: Subplots with a shared x and y axis
TIGHT LAYOUT
plt.tight_layout() adjusts subplot parameters so that
the subplots fit well in the Figure.
Examples:
If you do not use plt.tight_layout(), subplots might

overlap:
axes = axes.ravel()
ax.plot(series[i])
ax.set_title('Subplot ' + str(i))

…

diagram:
Figure 3.32: Subplots with no layout option
Using plt.tight_layout() results in no overlapping of

the subplots:
axes = axes.ravel()
ax.plot(series[i])
ax.set_title('Subplot ' + str(i))
plt.tight_layout()
…

diagram:
Figure 3.33: Subplots with a tight layout
RADAR CHARTS
Radar charts, also known as spider or web charts, visualize
multiple variables, with each variable plotted on its own axis,
resulting in a polygon. All axes are arranged radially, starting
at the center with equal distance between each other and have
the same scale.
EXERCISE 5: WORKING
ON RADAR CHARTS
In this exercise, it is shown step-by-step how to create a radar
chart:

this exercise.


Notebook:
# Import settings
import numpy as np
import pandas as pd
import
plt
%matplotlib inline
3. The following dataset contains

ratings of five different attributes
for four employees:
# Sample data
# Attributes:
Efficiency, Quality,
Commitment,
Responsible Conduct,
Cooperation
data = pd.DataFrame({
'Employee': ['A', 'B',

'C', 'D'],
'Efficiency': [5, 4,
4, 3,],
'Quality': [5, 5, 3,
3],
'Commitment': [5, 4,
4, 4],
'Responsible Conduct':
[4, 4, 4, 3],
'Cooperation': [4, 3,
4, 5]
})
4. Create angle values and close the

plot:
attributes =
list(data.columns[1:])
values =
list(data.values[:,
1:])
employees =
list(data.values[:,
0])
angles = [n /
float(len(attributes))
* 2 * np.pi for n in
range(len(attributes))
]
# Close the plot
angles += angles[:1]
values =
np.asarray(values)
values =
np.concatenate([values
, values[:, 0:1]],
axis=1)
5. Create subplots with the polar

projection. Set a tight layout so that
nothing overlaps:
# Create figure
8), dpi=150)
# Create subplots
for i in range(4):
ax = plt.subplot(2, 2,
i + 1, polar=True)
ax.plot(angles,
values[i])
ax.set_yticks([1, 2,
3, 4, 5])
ax.set_xticks(angles)
ax.set_xticklabels(att
ributes)
ax.set_title(employees
[i], fontsize=14,
color='r')
# Set tight layout
plt.tight_layout()
# Show plot
plt.show()
The following figure shows the output of the preceding code:

Figure 3.34: Radar charts
GRIDSPEC
matplotlib.gridspec.GridSpec(nrows, ncols)
specifies the geometry of the grid in which a subplot will be
placed.
Example:
…
gs = matplotlib.gridspec.GridSpec(3, 4)
ax1 = plt.subplot(gs[:3, :3])
ax2 = plt.subplot(gs[0, 3])
ax1.plot(series[0])
ax2.plot(series[1])
ax3.plot(series[2])
ax4.plot(series[3])
plt.tight_layout()

diagram:
Figure 3.35: GridSpec
ACTIVITY 18: CREATING

SCATTER PLOT WITH
MARGINAL HISTOGRAMS
In this activity, we will make use of GridSpec to visualize a
scatter plot with marginal histograms:
1. The given dataset, AnAge, which

was already used in the previous
activity, is not complete. Filter the
data so that you end up with
samples containing a body mass
and a maximum longevity. Select
all of the samples of the Aves class
with body mass less than 20,000.
2. Create a Figure with a constrained

layout. Create a gridspec of size
4x4. Create a scatter plot of size
3x3 and marginal histograms of
size 1x3 and 3x1. Add labels and a
Figure title.

follows:
Figure 3.36: Scattered plots with marginal
histograms
Note:
Images
In case you want to include images in your visualizations or in
case you are working with image data, Matplotlib offers
several functions to deal with images. In this section, we will
show you how to load, save, and plot images with Matplotlib.
Note
The images that are used in this topic are from
https://unsplash.com/.
BASIC IMAGE
OPERATIONS
Following are the basic operations that are used for designing
an image:
Loading Images
In case you encounter image formats that are not supported by

Matplotlib, we recommend using the Pillow library for loading
the image. In Matplotlib, loading images is part of the image
submodule. We use the alias mpimg for the submodule, as
follows:
import matplotlib.image as mpimg
mpimg.imread(fname) reads an image and returns it as a

numpy.array. For grayscale images, the returned array has
a shape (height, width), for RGB images (height, width, 3), and
for RGBA images (height, width, 4). The array values range
from 0 to 255.
Saving images
mpimg.imsave(fname, array) saves a

numpy.array as an image file. If the format parameter is
not given, the format is deduced from the filename extension.
With the optional parameters vmin and vmax, the color limits
can be set manually. For a grayscale image, the default for the
optional parameter cmap is 'viridis'; you might want to
change it to 'gray'.
Plotting a Single Image
plt.imshow(img) displays an image and returns an

AxesImage. For grayscale images with shape (height, width),
the image array is visualized using a colormap. The default
colormap is 'viridis', as illustrated in Figure 3.38. For
actually visualizing a grayscale image, the colormap has to be
set to 'gray', that is, plt.imshow(img,
cmap='gray'), which is illustrated in the following
diagram. Values for grayscale, RGB, and RGBA images can be
either float or uint8, and range from [0…1] or [0…
255], respectively. To manually define the value range, the
parameters vmin and vmax must be specified. A visualization
of an RGB image is shown in the following diagrams:
Figure 3.37: Grayscale image with default viridis

colormap
The following Figure shows a Grayscale image with gray

colormap:
Figure 3.38: Grayscale image with gray colormap
The following Figure shows a RGB image:

Figure 3.39: RGB image
Sometimes, it might be helpful to get an insight into the color

values. We can simply add a color bar to the image plot. It is
recommended to use a colormap with a high contrast, for
example, 'jet':
plt.imshow(img, cmap='jet')
plt.colorbar()
The preceding example is illustrated in the following diagram:

Figure 3.40: Image with jet colormap and color bar
Another way to get insight into the image values is to plot a

histogram, as shown in the following diagram. To plot the
histogram for an image array, the array has to be flattened
using numpy.ravel:
plt.hist(img.ravel(), bins=256, range=

(0, 1))
The following Figure shows an output of the preceding code:

Figure 3.41: Histogram of image values
Plotting Multiple Images in a Grid
For plotting multiple images in a grid, we can simply use

plt.subplots and plot an image per Axes:
for i in range(2):
axes[i].imshow(imgs[i])

diagram:
Figure 3.42: Multiple images within a grid
In some situations, it would be neat to remove the ticks and

add labels. axes.set_xticks([]) and
axes.set_yticks([]) remove x-ticks and y-ticks,
respectively. axes.set_xlabel('label') adds a label:
labels = ['coast', 'beach']
for i in range(2):
axes[i].imshow(imgs[i])
axes[i].set_xticks([])
axes[i].set_yticks([])
axes[i].set_xlabel(labels[i])
…

diagram:
Figure 3.43: Multiple images with labels

GRID
In this activity, we will plot images in a grid:
1. Load all four images from the

subfolder data.
2. Visualize the images in a 2x2 grid.
Remove the axes and give each
image a label.

follows:
Figure 3.44: Visualizing images in the 2x2 grid
Note:
Writing Mathematical
Expressions
In case you need to write mathematical expressions within the
code, Matplotlib supports TeX. You can use it in any text by
placing your mathematical expression in a pair of dollar signs.
There is no need to have TeX installed since Matplotlib comes
with its own parser.
An example of this is given in the following code:
plt.xlabel('$x$')
plt.ylabel('$\cos(x)$')
The following Figure shows an output of the preceding code:

Figure 3.45: Figure demonstrating mathematical
expressions
TeX examples:
'$\alpha_i>\beta_i$'
produces
'$\sum_{i=0}^\infty
x_i$' produces
'$\sqrt[3]{8}$' produces
'$\frac{3 - \frac{x}{2}}
{5}$' produces
Summary
In this chapter, we provided a detailed introduction to
Matplotlib, one of the most popular visualization libraries for
Python. We started off with the basics of pyplot and its
operations, and then followed up with a deep insight into the
numerous possibilities that helps to enrich visualizations with
text. Using practical examples, this chapter covered the most
popular plotting functions that Matplotlib offers out of the box,
including comparison charts, composition, and distribution
plots. This chapter is rounded off with how to visualize images
and write mathematical expressions.
In the next chapter, we will learn about the Seaborn library.

Seaborn is built on top of Matplotlib and provides a higher-
level abstraction to make visually appealing visualizations. We
will also address advanced visualization types.
Chapter 4
Simplifying
Seaborn
Learning Objectives
Explain why Seaborn is better than

Matplotlib
Design visually appealing plots

efficiently
Create insightful Figures
In this chapter, we will see how Seaborn is different than

Matplolib and also construct effective plots using Figure.
Introduction
Unlike Matplotlib, Seaborn is not a standalone Python library.
It is built on top of Matplotlib and provides a higher-level
abstraction to make visually appealing statistical visualizations.
A neat feature of Seaborn is the ability to integrate with
DataFrames from the pandas library.
With Seaborn, we attempt to make visualization a central part
of data exploration and understanding. Internally, Seaborn
operates on DataFrames and arrays that contain the complete
dataset. This enables it to perform semantic mappings and
statistical aggregations that are essential for displaying
informative visualizations. Seaborn can also be solely used to
change the style and appearance of Matplotlib visualizations.
The most prominent features of Seaborn are as follows:
Beautiful out of the box plots with

different themes
Built-in color palettes that can be

used to reveal patterns in the
dataset
Dataset-oriented interface
A high-level abstraction that still

allows for complex visualizations
ADVANTAGES OF
SEABORN
Seaborn is built on top of Matplotlib, and also addresses some
of the main pain points of working with Matplotlib.
Working with DataFrames using Matplotlib adds some

inconvenient overhead. For example: simply exploring your
dataset can take up a lot of time, since you need some
additional data wrangling to be able to plot the data from the
DataFrames using Matplotlib.
Seaborn, however, is built to operate on DataFrames and full

dataset arrays, which makes this process simpler. It internally
performs the necessary semantic mappings and statistical
aggregation to produce informative plots. The following is an
example of plotting using the Seaborn library:
import seaborn as sns
import pandas as pd
sns.set(style="ticks")
data = pd.read_csv("data/salary.csv")
sns.relplot(x="Salary", y="Age",
hue="Education", style="Education",
col="Gender", data=data)
This creates the following plot:

Figure 4.1: Seaborn Relation plot
Behind the scenes, Seaborn uses Matplotlib to draw plots.

Even though many tasks can be accomplished with just
Seaborn, a further customization might require the usage of
Matplotlib. We only provided the names of the variables in the
dataset and the roles they play in the plot. Unlike in Matplotlib,
it is not necessary to translate the variables into parameters of
the visualization.
Other pain points are the default Matplotlib parameters and

configurations. The default parameters in Seaborn provide
better visualizations without additional customization. We will
look at these default parameters in detail in the upcoming
topics.
For users who are already familiar with Matplotlib, the

extension with Seaborn is trivial, since the core concepts are
mostly similar.
Controlling Figure
Aesthetics
As we mentioned previously, Matplotlib is highly
customizable. But this also has the effect that it is difficult to
know what settings to tweak to achieve a visually appealing
plot. In contrast, Seaborn provides several customized themes
and a high-level interface for controlling the appearance of
Matplotlib figures.
The following code snippet creates a simple line plot in

Matplotlib:
%matplotlib inline
plt.figure()
x1 = [10, 20, 5, 40, 8]
x2 = [30, 43, 9, 7, 20]
plt.plot(x1, label='Group A')
plt.plot(x2, label='Group B')

plt.legend()
plt.show()
This is what the plot looks like with Matplotlib's default

parameters:
Figure 4.2: Matplotlib line plot
To switch to the Seaborn defaults, simply call the set()

function:
%matplotlib inline
sns.set()
plt.figure()
x1 = [10, 20, 5, 40, 8]
x2 = [30, 43, 9, 7, 20]
plt.legend()
plt.show()
The plot is as follows:

Figure 4.3: Seaborn line plot
Seaborn categorizes Matplotlib's parameters into two groups.

The first group contains parameters for the aesthetics of the
plot, while the second group scales various elements of the
figure so that it can be easily used in different contexts, such as
visualizations that are used for presentations, posters, and so
on.
SEABORN FIGURE
STYLES
To control the style, Seaborn provides two methods:
set_style(style, [rc]) and axes_style(style,
[rc]).
seaborn.set_style(style, [rc]) sets the aesthetic

style of the plots.
Parameters:
style: A dictionary of parameters

or the name of one of the following
preconfigured sets: darkgrid,
whitegrid, dark, white, or
ticks
rc (optional): Parameter mappings

to override the values in the preset
Seaborn style dictionaries
Here is an example:
%matplotlib inline
sns.set_style("whitegrid")
plt.figure()
x1 = [10, 20, 5, 40, 8]
x2 = [30, 43, 9, 7, 20]
plt.legend()
plt.show()
This results in the following plot:

Figure 4.4: Seaborn line plot with whitegrid style
seaborn.axes_style(style, [rc]) returns a

parameter dictionary for the aesthetic style of the plots. The
function can be used in a with statement to temporarily change
the style parameters.
Here are the parameters:
style: A dictionary of parameters

or the name of one of the following
preconfigured sets: darkgrid,
whitegrid, dark, white, or
ticks

Seaborn style dictionaries.
Here is an example:
%matplotlib inline
sns.set()
plt.figure()
x1 = [10, 20, 5, 40, 8]
x2 = [30, 43, 9, 7, 20]
with sns.axes_style('dark'):
plt.legend()
plt.show()
The aesthetics are only changed temporarily. The result is

Figure 4.5: Seaborn Line plot with dark axes style
For further customization, you can pass a dictionary of

parameters to the rc argument. You can only override
parameters that are part of the style definition.
REMOVING AXES SPINES

Sometimes, it might be desirable to remove the top and right
axes spines.
seaborn.despine(fig=None, ax=None,
top=True, right=True, left=False,
bottom=False, offset=None, trim=False)
removes the top and right spines from the plot.
The following code helps to remove the axes spines:
%matplotlib inline
sns.set_style("white")
plt.figure()
x1 = [10, 20, 5, 40, 8]
x2 = [30, 43, 9, 7, 20]
sns.despine()
plt.legend()
plt.show()
This results in the following plot:
Figure 4.6: Despined Seaborn line plot
CONTEXTS
A separate set of parameters controls the scale of plot elements.
This is a handy way to use the same code to create plots that
are suited for use in contexts where larger or smaller plots are
necessary. To control the context, two functions can be used.
seaborn.set_context(context, [font_scale],
[rc]) sets the plotting context parameters. This does not
change the overall style of the plot, but affects things such as
the size of the labels, lines, and so on. The base context is
notebook, and the other contexts are paper, talk, and
poster, which are versions of the notebook parameters
scaled by 0.8, 1.3, and 1.6, respectively.
context: A dictionary of
parameters or the name of one of
the following preconfigured sets:
paper, notebook, talk, or poster
font_scale (optional): A scaling

factor to independently scale the
size of font elements

Seaborn context dictionaries
The following code helps set the context:
%matplotlib inline
sns.set_context("poster")
plt.figure()
x1 = [10, 20, 5, 40, 8]
x2 = [30, 43, 9, 7, 20]
plt.legend()
plt.show()
The preceding code generates the following output:
Figure 4.7: Seaborn line plot with poster context
seaborn.plotting_context(context,
[font_scale], [rc]) returns a parameter dictionary to
scale elements of the Figure. This function can be used with a
statement to temporarily change the context parameters.
context: A dictionary of
parameters or the name of one of
the following preconfigured sets:
paper, notebook, talk, or poster
font_scale (optional): A scaling

factor to independently scale the
size of font elements

Seaborn context dictionaries
IQ SCORES FOR
DIFFERENT TEST
PLOT
In this activity, we will compare IQ scores among different test
groups using a box plot of the Seaborn library:

2. Access the data of each group in the

column, convert them into a list,
and assign this list to the variables
of each respective group.
3. Create a pandas DataFrame using

the preceding data by using the data

scores of different test groups using
Seaborn's boxplot function.
5. Use the whitegrid style, set the

context to talk, and remove all axes
spines, except the one on the
bottom. Add a title to the plot.
the final output should be as
follows:
Figure 4.8: IQ scores of groups
Note:
Color Palettes
Color is a very important factor for your visualization. Color
can reveal patterns in the data if used effectively or hide
patterns if used poorly. Seaborn makes it easy to select and use
color palettes that are suited to your task. The
color_palette() function provides an interface for many
of the possible ways to generate colors.
seaborn.color_palette([palette],
[n_colors], [desat]) returns a list of colors, thus
defining a color palette.
palette (optional): Name of

palette or None to return the current
palette.
n_colors (optional): Number of

colors in the palette. If the specified
number of colors is larger than the
number of colors in the palette, the
colors will be cycled.
desat (optional): Proportion to

desaturate each color by.
You can set the palette for all plots with set_palette().
This function accepts the same arguments as
color_palette(). In the following sections, we will
explain how color palettes are divided into different groups.
CATEGORICAL COLOR
PALETTES
Categorical palettes are best for distinguishing discrete data
that does not have an inherent ordering. There are six default
themes in Seaborn: deep, muted, bright, pastel, dark,
and colorblind. The code and output for each theme is
provided in the following code:
palette1 = sns.color_palette("deep")
sns.palplot(palette1)
Figure 4.9: Deep color palette
palette2 = sns.color_palette("muted")
Figure 4.10: Muted color palette
The following figure shows a bright color palette:
palette3 = sns.color_palette("bright")
Figure 4.11: Bright color palette
The following figure shows a pastel color palette:
palette4 = sns.color_palette("pastel")
Figure 4.12: Pastel color palette

The following figure shows a dark color palette:
palette5 = sns.color_palette("dark")
Figure 4.13: Dark color palette
palette6 =
sns.color_palette("colorblind")
Figure 4.14: Colorblind color palette
SEQUENTIAL COLOR
PALETTES
Sequential color palettes are appropriate when the data ranges
from relatively low or uninteresting values to relatively high or
interesting values. The following code snippets, along with
their respective outputs, give us a better insight into sequential
color palettes:
custom_palette2 =
sns.light_palette("brown")
sns.palplot(custom_palette2)

Figure 4.15: Custom brown color palette
The preceding palette can also be reversed by setting the

reverse parameter to True in the following code:
custom_palette3 =
sns.light_palette("brown", reverse=True)
Figure 4.16: Custom reversed brown color palette
DIVERGING COLOR
PALETTES
Diverging color palettes are used for data that
consists of a well-defined midpoint. Emphasis is being laid on
both high and low values. For example: if you are plotting any
population changes for a particular region from some baseline
population, it is best to use diverging colormaps to show the
relative increase and decrease in the population. The following
code snippet and output provides a better understanding about
diverging plot, wherein we use the coolwarm template,
which is built into Matplotlib:
custom_palette4 =
sns.color_palette("coolwarm", 7)

Figure 4.17: Coolwarm color palette
You can use the diverging_palette() function to create

custom diverging palettes. We can pass two hues in degrees as
parameters, along with the total number of palettes. The
following code snippet and output provides better insight:
custom_palette5 =
sns.diverging_palette(440, 40, n=7)
Figure 4.18: Custom diverging color palette
ACTIVITY 21: USING

HEATMAPS TO FIND
PATTERNS IN FLIGHT
PASSENGERS' DATA
In this activity, we will use a heatmap to find the patterns in the
flight passengers' data:

in the subfolder data. The given
dataset contains the monthly figures
for flight passengers from the years
2001 to 2012.
2. Use a heatmap to visualize the

given data.
3. Use your own color map. Make
sure that the lowest value is the
darkest color and that the highest
value is the brightest color.

follows:
Figure 4.19: Heatmap of Flight Passengers data
Note:
Interesting Plots in
Seaborn
In the previous chapter, we discussed various plots in
Matplotlib, but there are still a few visualizations left that we
want to discuss.
BAR PLOTS
In the last chapter, we already explained how to create bar plots
with Matplotlib. Creating bar plots with subgroups was quite
tedious, but Seaborn offers a very convenient way to create
various bar plots. They can be also used in Seaborn to
represent estimates of central tendency with the height of each
rectangle and indicates the uncertainty around that estimate
using error bars.
The following example gives you a good idea of how this

works:
import pandas as pd
sns.set(style="whitegrid")
sns.barplot(x="Education", y="Salary",
hue="District", data=data)
The result is shown in the following diagram:

Figure 4.20: Seaborn bar plot
ACTIVITY 22: MOVIE

In this activity, we will use a bar plot to compare movie scores.
You will be given five movies with scores from Rotten
Tomatoes. The Tomatometer is the percentage of approved
Tomatometer critics who have given a positive review for the
movie. The Audience Score is the percentage of users who
have given a score of 3.5 or higher out of 5. Compare these
two scores among the five movies:

2. Transform the data into a useable

format for Seaborn's bar plot
function.
3. Use Seaborn to create a visually

appealing bar plot that compares
the two scores for all five movies.

the expected output should look as
follows:
Figure 4.21: Movie Scores comparison
Note:
KERNEL DENSITY
ESTIMATION
It is often useful to visualize how variables of a dataset are
distributed. Seaborn offers handy functions to examine
univariate and bivariate distributions. One possible way to look
at a univariate distribution in Seaborn is by using the
distplot() function. This will draw a histogram and fit a
kernel density estimate (KDE), as illustrated in the following
example:
%matplotlib inline
import numpy as np
import pandas as pd
x = np.random.normal(size=50)
sns.distplot(x)
Figure 4.22: KDE with histogram for univariate

distribution
To just visualize the KDE, Seaborn provides the kdeplot()

function:
sns.kdeplot(x, shade=True)
The KDE plot is shown in the following diagram, along with a

shaded area under the curve:
Figure 4.23: KDE for univariate distribution
PLOTTING BIVARIATE
DISTRIBUTIONS
For visualizing bivariate distributions, we will introduce
three different plots. The first two plots use the
jointplot() function, which creates a multi-panel figure
that shows both the joint relationship between both variables
and the corresponding marginal distributions.
A scatter plot shows each observation as points on the x and y

axis. Additionally, a histogram for each variable is shown:
import pandas as pd
sns.set(style="white")
sns.jointplot(x="Salary", y="Age",
data=data)
The scatter plot with marginal histograms is shown in the

following diagram:
Figure 4.24: Scatter plot with marginal histograms
It is also possible to use the KDE procedure to visualize

bivariate distributions. The joint distribution is shown as a
contour plot, as demonstrated in the following code:
sns.jointplot('Salary', 'Age',
data=subdata, kind='kde', xlim=(0,
500000), ylim=(0, 100))

Figure 4.25: Contour plot
VISUALIZING PAIRWISE
RELATIONSHIPS
For visualizing multiple pairwise bivariate distributions in a
dataset, Seaborn offers the pairplot() function. This
function creates a matrix where off-diagonal elements visualize
the relationship between each pair of variables and the
diagonal elements show the marginal distributions.
The following example gives us a better understanding of this:
%matplotlib inline
import numpy as np
import pandas as pd
mydata =
pd.read_csv("data/basic_details.csv")
sns.set(style="ticks", color_codes=True)
g = sns.pairplot(mydata, hue="Groups")
The pair plot, also called a correlogram, is shown in the

following diagram. Scatter-plots are shown for all variable
pairs on the off-diagonal, while KDEs are shown on the
diagonal. Groups are highlighted by different colors:
Figure 4.26: Seaborn pair plot
VIOLIN PLOTS
A different approach to visualizing statistical measures is by
using violin plots. They combine box plots with the kernel
density estimation procedure that we described previously. It
provides a richer description of the variable's distribution.
Additionally, the quartile and whisker values from the box plot
are shown inside the violin.
The following example demonstrates the usage of violin plots:
import pandas as pd

sns.set(style="whitegrid")
sns.violinplot('Education', 'Salary',
hue='Gender', data=data, split=True,
cut=0)
The result looks as follows:

Figure 4.27: Seaborn violin plot

IQ SCORES FOR
DIFFERENT TEST
GROUPS BY USING A
VIOLIN PLOT
In this activity, we will compare IQ scores among different test
groups by using the violin plot that's provided by Seaborn's
library:


and assign this list to the variables


scores of different test groups using
Seaborn's violinplot function.
5. Use the whitegrid style, set the

context to talk, and remove all axes
spines, except the one on the
bottom. Add a title to the plot.

the final output should look as
follows:
Figure 4.28: Violin plot showing IQ scores of different

groups
Note:
Multi-Plots in Seaborn
In the previous topic, we introduced a multi-plot, namely the
pair plot. In this topic, we want to talk about a different way to
create flexible multi-plots.
FACETGRID
The FacetGrid is useful for visualizing a certain plot for
multiple variables separately. A FacetGrid can be drawn with
up to three dimensions: row, col, and hue. The first two
have the obvious correspondence to the rows and columns of
an array. The hue is the third dimension and shown with
different colors. The FacetGrid class has to be initialized
with a DataFrame and the names of the variables that will form
the row, column, or hue dimensions of the grid. These variables
should be categorical or discrete.
seaborn.FacetGrid(data, row, col, hue, …)

initializes a multi-plot grid for plotting conditional
relationships.
Here are some interesting parameters:
data: A tidy ("long-form")

DataFrame where each column
corresponds to a variable and each
row corresponds to an observation
row, col, hue: Variables that

define subsets of the given data,
which will be drawn on separate
facets in the grid
sharex, sharey (optional):

Share x/y axes across
rows/columns
height (optional): Height (in
inches) of each facet
Initializing the grid does not draw anything on them yet. For
visualizing data on this grid, the FacetGrid.map() method
has to be used. You can provide any plotting function and the
name(s) of the variable(s) in the data frame to plot.
FacetGrid.map(func, *args, **kwargs) applies a

plotting function to each facet of the grid.
func: A plotting function that

takes data and keyword arguments.
*args: The column names in data

that identify variables to plot. The
data for each variable is passed to
func in the order the variables are
specified in.
**kwargs: Keyword arguments

that are passed to the plotting
function.
The following example visualizes FacetGrid with scatter plot:
import pandas as pd
g = sns.FacetGrid(subdata,
col='District')
g.map(plt.scatter, 'Salary', 'Age')
Figure 4.29: FacetGrid with scatter plots
ACTIVITY 24: TOP 30

YOUTUBE CHANNELS
In this activity, we will visualize the total number of
subscribers and the total number of views for the top 30
YouTube channels by using the FacetGrid() function that's
provided by the Seaborn library. Visualize the given data using
a FacetGrid with two columns. The first column should show
the number of subscribers for each YouTube channel, whereas
the second column should show the number of views.
Following are the steps to implement this activity:


and assign this list to variables of
each respective group.

4. Create a FacetGrid with two

columns to visualize the data.

the final output should look as
follows:
Figure 4.30: Subscribers and Views of top 30
YouTube channels
Note:
Regression Plots
Many datasets contain multiple quantitative variables, and the
goal is to find a relationship among those variables. We
previously mentioned a few functions that show the joint
distribution of two variables. It can be helpful to estimate
relationships between two variables. We will only cover linear
regression in this topic; however, Seaborn provides a wider
range of regression functionality if needed.
To visualize linear relationships, determined through linear
regression, the regplot() function is offered by Seaborn.
The following code snippet gives a simple example:
import numpy as np
x = np.arange(100)
y = x + np.random.normal(0, 5, size=100)
sns.regplot(x, y)
The regplot() function draws a scatter plot, a regression

line, and a 95% confidence interval for that regression, as
Figure 4.31: Seaborn regression plot
ACTIVITY 25: LINEAR

REGRESSION
In this activity, we will use a regression plot to visualize the
linear relationships:

2. Filter the data so that you end up

Only consider samples for the
Mammalia class and a body mass
below 200,000.
3. Create a regression plot to visualize

the linear relationship of the
variables.

the output should look as follows:
Figure 4.32: Linear regression for animal attribute
relations
Note:
Squarify
At this point, we will briefly talk about tree maps. Tree maps
display hierarchical data as a set of nested rectangles. Each
group is represented by a rectangle, of which its area is
proportional to its value. Using color schemes, it is possible to
represent hierarchies: groups, subgroups, and so on. Compared
to pie charts, tree maps efficiently use space. Matplotlib and
Seaborn do not offer tree maps, and so the Squarify library
that is built on top of Matplotlib is used. Seaborn is a great
addition for creating color palettes.
The following code snippet is a basic tree map example. It
requires the Squarify library:
%matplotlib inline
import squarify
colors = sns.light_palette("brown", 4)
squarify.plot(sizes=[50, 25, 10, 15],

label=["Group A", "Group B", "Group C",
"Group D"], color=colors)
plt.axis("off")
plt.show()
Figure 4.33: Tree map
ACTIVITY 26: WATER

USAGE REVISITED
In this activity, we will use a tree map to visualize the
percentage of water that's used for different purposes:

2. Use a tree map to visualize the

water usage.
3. Show the percentages for each tile

and add a title.

follows:
Figure 4.34: Tree map of water usage
Note:
Summary
In this chapter, we demonstrated how Seaborn helps create
visually appealing figures. We discussed various options for
controlling figure aesthetics, such as figure style, controlling
spines, and setting the context of visualizations. We talked
about color palettes in detail. Further visualizations were
introduced for visualizing univariate and bivariate
distributions. Moreover, we discussed FacetGrids, which can
be used for creating multi-plots, and regression plots as a way
to analyze the relationships between two variables. Finally, we
discussed the Squarify library, which is used to create tree
maps. In the next chapter, we will show you how to visualize
geospatial data in various ways by using the Geoplotlib library.
Chapter 5
Plotting Geospatial
Data
Learning Objectives
Utilize Geoplotlib to create

stunning geographical
visualizations
Identify the different types of

geospatial charts
Demonstrate datasets containing

geospatial data for plotting
Explain the importance of plotting

geospatial information
In this chapter, we will use the geoplotib library to visualize

different geospatial data.
Introduction
Geoplotlib is an open source Python library for geospatial data
visualizations. It has a wide range of geographical
visualizations and supports hardware acceleration. It also
provides performance rendering for large datasets with millions
of data points. As discussed in the earlier chapters, Matplotlib
provides the ways to visualize geographical data. However,
Matplotlib is not designed for this task, as its interfaces are
complex and inconvenient to use. Matplotlib also restricts the
ways in which geographical data can be displayed. The
Basemap and Cartopy libraries enable it so that you can plot
on a world map. However, these packages do not support
drawing on map tiles.
Geoplotlib, on the other hand, was designed exactly for this

purpose and therefore it not only provides map tiles but also
allows for interactivity and simple animations. It provides a
simple interface that allows access to powerful geospatial
visualizations.
Note
For a better understanding of the available features of
Geoplotlib, you can visit the following link:
https://github.com/andrea-cuttone/geoplotlib/wiki/User-Guide.
To understand the concepts, design, and implementation of

Geoplotlib, let's take a brief look at its conceptual architecture.
The two inputs that are fed to Geoplotlib are your data sources
and the map tiles. The map tiles, as we'll see later, can be
replaced by different providers. The outputs describe the
possibilities to not only render images inside the Jupyter
Notebooks but also to work in an interactive window that
allows for zooming and panning the maps. The schema of the
components of Geoplotlib looks like this:
Figure 5.1: Conceptual architecture of geoplotlib
Geoplotlib uses the concept of layers that can be placed on top

one another, providing a powerful interface for even complex
visualizations. It comes with several common visualization
layers that are easy to set up and use.
From the preceding diagram, we can see that geoplotlib is built

on top of NumPy/SciPy, and Pyglet/OpenGL. These libraries
take care of numerical operations and rendering. Both
components are based on Python, thus enabling the usage of
the full Python ecosystem.
THE DESIGN PRINCIPLES

OF GEOPLOTLIB
Taking a closer look at the internal design of Geoplotlib, we
can see that it is built around three design principles:
Simplicity: Looking at the example

provided here, we can quickly see
that Geoplotlib abstracts away the
complexity of plotting map tiles
and the already provided layers
such as dot-density and histogram.
It has a simple API that provides
common visualizations. These
visualizations can be created using
custom data with only a few lines
of code. If our dataset comes with
lat and lon columns, we can display
those datapoints as dots on a map
with five lines of code, like this:
import geoplotlib
from geoplotlib.utils
import read_csv
dataset =
read_csv('./data/poach
ing_points_cleaned.csv
')
geoplotlib.dot(dataset
)
geoplotlib.show()
In addition to this, everyone who's

used Matplotlib before will have no
problems understanding the syntax
of Geoplotlib, because its syntax is
inspired by the syntax of
Matplotlib.
Integration: Geoplotlib
visualizations are purely Python-
based. This means that the generic
Python code can be executed and
other libraries such as pandas can
be used for data wrangling
purposes. We can manipulate and
enrich our datasets using pandas
DataFrames and later simply
convert them into a Geoplotlib
DataAccessObject, which we
need for optimum compatibility,
like this:
import pandas as pd
import
DataAccessObject
pd_dataset =
pd.read_csv('./data/po
aching_points_cleaned.
csv')
# data wrangling with

pandas DataFrames here
dataset =
DataAccessObject(pd_da
taset)
Geoplotlib fully integrates into the
Python ecosystem. This enables us
to even plot geographical data
inline inside our Jupyter
Notebooks. This possibility allows
us to design our visualizations
quickly and iteratively.
Performance: As we mentioned
before, Geoplotlib is able to handle
large amounts of data due to the
usage of NumPy for accelerated
numerical operations and OpenGL
for accelerated graphical rendering.
GEOSPATIAL
VISUALIZATIONS
Choropleth plot, Voronoi tessellation, and Delaunay
triangulation are a few of the geospatial visualizations that
will be used in this chapter. The explanation for each of them is
provided here:
Choropleth Plot
This kind of geographical plot displays areas such as states of a

country in a shaded or colored manner. The shade or color is
determined by a single or set of data points. It gives an abstract
view of a geographical area to visualize the relations and
differences between the different areas. In Figure 5.21, we can
see that the shade of each state of the US is determined by the
percentage of obesity. The darker the shade, the higher the
percentage.
Voronoi tessellation
In a Voronoi tessellation, each pair of data points is basically

separated by a line that has the same distance from both data
points. The separation creates cells that, for every given point,
marks which data point is closer. The closer the data points, the
smaller the cells.
Delaunay triangulation
A Delaunay triangulation is related to the Voronoi

tessellation. When connecting each data point to every other
data point that shares an edge, we end up with a plot that is
triangulated. The closer the data points, the smaller the
triangles will be. This gives us a visual clue about the density
of points in specific areas. When combined with color
gradients, we get insights about points of interests, which can
be compared with a heatmap.
EXERCISE 6: VISUALIZING
SIMPLE GEOSPATIAL
DATA
In this exercise, we'll be looking at the basic usage of
Geoplotlib's plot methods for DotDensity, Histograms, and
Voronoi diagrams. For this, we will make use of the data of
various poaching incidents that have taken place all over the
world:

this exercise.

the path of this file. In the
command-line terminal, type:
jupyter-lab
2. By now, you should be familiar

with the process of working with
Jupyter Notebooks.

file.
4. As always, first, we want to import

the dependencies that we will need.
In this case, we will work without
pandas, since geoplotlib has its
own read_csv method that
enables us to read a .csv file into a
DataAccessObject:
# importing the
import geoplotlib
import read_csv
5. The data is being loaded in the

same way as with the pandas
read_csv method:
dataset =
read_csv('./data/poach
ing_points_cleaned.csv
')
Note
The preceding dataset can be found

here: https://bit.ly/2Xosg2b.
6. The dataset is stored in a

DataAccessObject class that's
provided by Geoplotlib. It does not
have the same capabilities as
pandas DataFrames. It's meant for
the simple and quick loading of
data so that you can create a
visualization. If we print out this
object, we can see the difference
better. It gives us a basic overview
of what columns are present and
how many rows the dataset has:
# looking at the
dataset structure
Dataset
The following figure shows the

output of the preceding code:
Figure 5.2: Dataset structure
As we can see in the preceding

screenshot, the dataset consists of
268 rows and six columns. Each
row is uniquely identified by
id_report. The date_report
column states on what date the
poaching incident took place. On
the other hand, the
created_date column states the
date on which the report was
created. The description column
provides the basic information
about the incident. The lat and
lon columns state the geographical
location of the place where
poaching took place.
7. Geoplotlib is compatible with

pandas DataFrames as well. If you
need to do some pre-processing
with the data, it might make sense
to use pandas right away:
# csv import with

pandas
import pandas as pd
pd_dataset =
pd.read_csv('./data/po
aching_points_cleaned.
csv')
pd_dataset.head()
Figure 5.3: First five entries of the

dataset
Note
Geoplotlib requires your dataset to

have lat and lon columns. These
columns are the geographical data
for latitude and longitude, and are
used to determine how to plot.
8. To start with, we'll be using a

simple DotDensityLayer that will
plot each row of our dataset as a
single point on a map. Geoplotlib
comes with the dot method, which
creates this visualization without
further configurations:
Note
After setting up DotDensityLayer

here, we need to call the show
method, which will render the map
with the given layer.
# plotting our dataset

with points
)
geoplotlib.show()

Figure 5.4: Dot-density visualization of

poaching points
Only looking at the lat and lon

values in the dataset won't give us a
very good idea. We're not able to
draw conclusions and get insights
into our dataset without visualizing
our data points on a map. When
looking at the rendered map, we
can see that there are some more
popular spots with more dots and
some less popular ones with fewer
dots.
9. We now want to look at the point

density some more. To better
visualize the density, we have a few
options. One of them is to use a
histogram plot. We can define a
binsize, which will allow us to
set the size of the hist bins in our
visualization. Geoplotlib provides
the hist method, which will
create a Histogram Layer on top
of our map tiles:
# plotting our dataset

as a histogram
geoplotlib.hist(datase
t, binsize=20)
geoplotlib.show()

Figure 5.5: Histogram visualization of
poaching points
Histogram plots give us a better

understanding of the density
distribution for our dataset.
Looking at the final plot, we can
see that there are some hotspots for
poaching. This also highlights the
areas without any poaching
incidents.
10. Voronoi plots are also good for

visualizing the density of data
points. Voronoi introduces a little
bit more complexity with several
parameters such as cmap,
max_area, and alpha. Here,
cmap denotes the color of the map,
alpha denotes the color of the
alpha, and max_area denotes a
constant that determines the color
of the Voronoi areas. They are
useful if you want to better fit your
visualization into the data:
# plotting a voronoi
map
geoplotlib.voronoi(dat
aset, cmap='Blues_r',
max_area=1e5,
alpha=255)
geoplotlib.show()

Figure 5.6: Voronoi visualization of poaching points
If we compare this Voronoi visualization with the histogram

plot, we can see that one area that draws a lot of attention. The
center right edge of the plot shows quite a large dark blue area
with an even darker center, something that could've easily been
overlooked with the histogram plot.
Congratulations! We just covered the basics of Geoplotlib. It

has many more methods, but they all have a similar API that
makes using the other methods simple. Since we looked at
some of the very basic visualizations, it's now up to you to
solve the first activity.

MAP
In this activity, we will take the previously learned skills of
plotting data with Geoplotlib and apply them to the task of
finding the dense areas within the cities in Europe that have a
population of more than 100,000:

this activity.
2. Before we can start working with

the data, we will need to import the
dependencies.
3. Load the dataset using pandas.
4. After loading the dataset, we want

to list all the datatypes that are
present in it.
5. Once we can see that our datatypes

are correct, we will map our
Latitude and Longitude
columns to lat and lon.
6. Now, plot the data points in a dot

plot.
7. To get one step closer to solving

our given task, we need to get the
number of cities per country (the
first 20 entries) and filter out the
countries with a population greater
than zero.
8. Plot the remaining data in a dot
plot.
9. Again, filter your remaining data

for cities with a population greater
than 100,000.
10. To get a better understanding of the

density of our data points on the
map, we want to use a Voronoi
tessellation layer.
11. Filter down the data even further to

only cities in countries such as
Germany and Great Britain.
12. Finally, use a Delaunay

triangulation layer to find the
most densely populated areas.
13. Observe the expected output of dot

plot:
Figure 5.7: Dot-density visualization of the reduced
dataset
The following is the expected output of the Voronoi plot:
Figure 5.8: Voronoi visualization of densely

populated cities
The following is the expected output of the Delaunay

triangulation:
Figure 5.9: Delaunay triangle visualization of cities in
Germany and Great Britain
Note:
EXERCISE 7:
CHOROPLETH PLOT WITH
GEOJSON DATA
In this exercise, we not only want to work with GeoJSON
data, but also want to see how we can create a choropleth
visualization. They are especially useful for displaying
statistical variables in shaded areas. In our case, the areas will
be the outlines of the states of the USA. Let's create a
choropleth visualization with the given GeoJSON data:
this exercise.
2. Load the dependencies for this

exercise:
# importing the
import json
import geoplotlib
from geoplotlib.colors
import ColorMap
import BoundingBox
3. Before we create the actual

visualization, we need to
understand the outlines of our
dataset. Since the geojson
method of Geoplotlib only needs a
path to the dataset instead of a
DataFrame or object, we don't need
to load it. However, since we still
want to see what kind of data we
are handling, we must open the
GeoJSON file and load it as a
json object. Given that, we can
then access its members by simple
indexing:
# displaying one of
the entries for the
states
with
open('data/National_Ob
esity_By_State.geojson
') as data:
dataset =
json.load(data)
first_state =
dataset.get('features'
)[0]
# only showing one

coordinate instead of
all points
first_state['geometry'
]['coordinates'] =
first_state['geometry'
]['coordinates'][0][0]
print(json.dumps(first
_state, indent=4))
4. The following output describes one

of the features that displays the
general structure of a GeoJSON
file. The properties of interest for us
are the NAME, Obesity, and the
geometry coordinates:
Figure 5.10: General structure of the

geojson file
Note
Geospatial applications prefer

GeoJSON files for persisting and
exchanging geographical data.
5. Depending on the information

present in the GeoJSON file, we
might need to extract some of it for
later mappings. For the obesity
database, we want to extract the
names of all the states of the US.
The following code does the same:
# listing the states

in the dataset
with
open('data/National_Ob
esity_By_State.geojson
') as data:
dataset =
json.load(data)
states =
[feature['properties']
['NAME'] for feature
in
dataset.get('features'
)]
print(states)

Figure 5.11: List of all cities in the

United States
6. If your GeoJSON file is valid,

meaning that it has the expected
structure, you can then use the
geojson method of Geoplotlib.
By only providing the path to the
file, it will plot the coordinates for
each feature in a blue color by
default:
# plotting the
information from the
geojson file
geoplotlib.geojson('da
ta/National_Obesity_By
_State.geojson')
geoplotlib.show()
After calling the show method, the

map will show up with a focus on
North America. In the following
diagram, we can already see the
borders of each state:
Figure 5.12: Map with outlines of the

states plotted
7. To assign a color that represents the

obesity of each state, we have to
provide the color parameter to
the geojson method. We can
provide different types depending
on the use case. Rather than
assigning a single value to each
state, we want the darkness to
represent the percentage of obese
people. To do this, we have to
provide a method for the color
property. Our method simply maps
the Obesity attribute to a
ColorMap class object with
enough levels to allow good
distinction:
# converting the
obesity into a color
cmap =
ColorMap('Reds',
alpha=255, levels=40)
def
get_color(properties):
return
cmap.to_color(properti
es['Obesity'],
maxvalue=40,scale='lin
')
8. We then provide the color mapping

to our color parameter. This,
however, won't fill the areas. We
therefore also have to set the fill
parameter to True. In addition, we
also want to keep the outlines of
our states visible. Here, we can
leverage the concept that
Geoplotlib is layer-based, so we
can simply call the same method
again, providing a white color and
setting the fill parameter to
false. We also want to make sure
that our view is displaying the
country USA. To do this we, again,
use one of the constants that are
provided by Geoplotlib:
# plotting the shaded

states and adding
another layer which
plots the state
outlines in white
# our BoundingBox
should focus the USA
_State.geojson',
fill=True,
color=get_color)
_State.geojson',
fill=False, color=
[255, 255, 255, 255])
geoplotlib.set_bbox(Bo
undingBox.USA)
geoplotlib.show()

the expected output is as follows:
Figure 5.13: Choropleth visualization showing

obesity in different states
A new window will open, displaying the country USA with the
areas of its states filled with different shades of red. The darker
areas represent higher obesity percentages.
Note
To give the user some more information for this plot, we could
also use the f_tooltip parameter to provide a tooltip for
each state, thus displaying the name and the percentage of
obese people.
Congratulations! You've already built different plots and

visualizations using Geoplotlib. In this exercise, we looked at
displaying data from a GeoJSON file and creating a
choropleth plot.
In the following topics, we will cover more advanced

customizations that will give you the tools to create more
powerful visualizations.
Tile Providers
Geoplotlib supports the usage of different tile providers. This
means that any OpenStreetMap tile server can be used as a
backdrop to our visualization. Some of the popular free tile
providers are Stamen Watercolor, Stamen Toner, Stamen
Toner Lite, and DarkMatter.
Changing the tile provider can be done in two ways:
Make use of built-in tile

providers
Geoplotlib contains a few built-in

tile providers with shortcuts. The
following code shows how to use it:
geoplotlib.tiles_provi
der('darkmatter')
Provide a custom object to the

tiles_provider method
By providing a custom object to

Geoplotlib's
tiles_provider() method,
you will not only get the access to
the url from which the map tiles
are being loaded, but also see the
attribution displayed in the
lower right corner of the
visualization. We are also able to
set a distinct caching directory for
the downloaded tiles. The following
code shows how to provide a
custom object:
der({
'url': lambda zoom,

xtile, ytile:
'http://a.tile.stamen.
com/watercolor/%d/%d/%
d.png' % (zoom, xtile,
ytile),
'tiles_dir':
'tiles_dir',
'attribution': 'Python
Data Visualization |
Packt'
})
The caching in tiles_dir is

mandatory, since, each time the
map is scrolled or zoomed into, we
query new map tiles if they are not
already downloaded. This can lead
to the tile provider refusing your
request due to many requests in a
short period of time.
In the following exercise, we'll have a quick look at how to

switch the map tile provider. It might not seem powerful at
first, but it can take your visualizations to the next level if
leveraged correctly.
EXERCISE 8: VISUALLY
COMPARING DIFFERENT
TILE PROVIDERS
This quick exercise will teach you how to switch the map tile
provider for your visualizations. Geoplotlib provides mappings
for some of the available and most popular map tiles. However,
we can also provide a custom object that contains the url of
some tile providers:

this exercise.

the path of this file and in the
command-line terminal type:
jupyter-lab
2. We won't use any dataset in this

exercise, since we want to focus on
the map tiles and tile providers.
Therefore, the only import we need
to do is geoplotlib itself:
# importing the
import geoplotlib
3. We know that Geoplotlib has a

layers approach to plotting. This
means that we can simply display
the map tiles without adding any
plotting layer on top:
# displaying the map

with the default tile
provider
geoplotlib.show()

Figure 5.14: World map with default tile
provider
This will display a completely

empty world map, since we haven't
specified any tile provider. By
default, it will use the CartoDB
Positron map tiles.
4. Geoplotlib provides several

shorthand accessors to common
map tile providers. The
tiles_provider method
allows us to simply provide the
name of the provider:
# using map tiles from

the dark matter tile
provider
der('darkmatter')
geoplotlib.show()

Figure 5.15: World map with darkmatter

map tiles
In this example, we used the

darkmatter map tiles. As you
can see, those are very dark and
will make your visualizations pop
out.
Note
We can also use different map tiles

such as watercolor, toner, toner-
lite, and positron in a similar way.
5. When using tile providers that are

not covered by geoplotlib, we
can pass a custom object to the
tiles_provider method. It
maps the current viewport
information to the url. The
tiles_dir parameter defines
where the tiles should be cached.
When changing the url, you also
have to change tiles_dir to see
the changes immediately. The
attribution gives you the
option to display custom text in the
right lower corner:
# using custom object

to set up tile
provider
der({
'url': lambda zoom,

xtile, ytile:
'http://a.tile.openstr
eetmap.fr/hot/%d/%d/%d
.png' % (zoom, xtile,
ytile),
'tiles_dir':
'custom_tiles',
'attribution': 'Custom
Tiles Provider -
Humanitarian map style
| Packt Courseware'
})
geoplotlib.show()

Figure 5.16: Humanitarian map tiles from custom tile

providers object
Some map tile providers have strict request limits, so it might

lead to warning messages if you're zooming too fast.
Congratulations! You now know how to change the tile

provider to give your visualization one more layer of
customizability. This also introduces us to another layer of
complexity. It all depends on the concept of our final product
and whether we want to use the "default" map tiles or some
artistic map tiles.
The next topic will cover how to create custom layers that can
go far beyond the ones we have described in this book. We'll
look at the basic structure of the BaseLayer class and what it
takes to create a custom layer.
Custom Layers
Now that we have covered the basics of visualizing geospatial
data with the built-in layers, and the methods to change the tile
provider, we will now focus on defining our own custom
layers. Custom layers allow you to create more complex data
visualizations. They also help with adding more interactivity
and animation to them. Creating a custom layer starts by
defining a new class that extends the BaseLayer class that's
provided by Geoplotlib. Besides the __init__ method that
initializes the class level variables, we also have to at least
extend the draw method of the already provided BaseLayer
class.
Depending on the nature of your visualization, you might also

want to implement the invalidate method, which takes care of
map projection changes such as zooming into your
visualization. Both the draw and invalidate methods
receive a Projection object that takes care of the latitude
and longitude mapping on our two-dimensional viewport.
These mapped points can be handed to an instance of a
BatchPainter object that provides primitives such as
points, lines, and shapes to draw those coordinates onto your
map.
Note
Since Geoplotlib operates on OpenGL, this process is highly
performant and can even draw complex visualizations quickly.
For more examples on how to create custom layers, visit the
following GitHub repository of Geoplotlib:
https://github.com/andrea-
cuttone/geoplotlib/tree/master/examples.

WITH CUSTOM LAYERS
In this activity, we will take a look at creating a custom layer
that will allow you to not only display geospatial data but also
animate your data points over time. We'll get a deeper
understanding of how Geoplotlib works and how the layers are
created and drawn. Our dataset does not only contain spatial
but also temporal information, which enables us to plot the
flights over time on our map:

this activity.
2. First, make sure to import the

necessary dependencies.
3. Load the
flight_tracking.csv dataset
using pandas.
4. Have a look at the dataset and its

features.
5. Since our dataset has columns that

are named Latitude and
Longitude instead of lat and
lon, rename those columns to their
short versions.
6. Our custom layer will animate the

flight data, which means that we
need to work with the timestamp
of our data. Date and time are two
separate columns, so we need to
merge these columns. Use the
provided to_epoch method and
create a new timestamp column.
7. Create a new TrackLayer that

extends Geoplotlib's BaseLayer.
8. Implement the methods

__init__, draw, and bbox for
the TrackLayer. Use the
provided BoundingBox to focus
on Leeds when calling the
TrackLayer.

the expected output is as follows:
Figure 5.17: Flight tracking visualization focused on

Leeds
Note:
Summary
In this chapter, we covered the basic and advanced concepts
and methods of Geoplotlib. It gave us a quick insight into the
internal processes and how to practically apply the library to
our own problem statements. Most of the time, the built-in
plots should suit your needs pretty well. Once you're interested
in having animated or even interactive visualizations, you will
have to create custom layers that enable those features.
In the following chapter, we'll get some hands-on experience

with the Bokeh library and build visualizations that can easily
be integrated into web pages. Once we have finished using
Bokeh, we'll conclude this chapterware with a chapter that
gives you the opportunity to work with a new dataset and a
library of your choice so that you can come up with your very
own visualization. This will be the last step in consolidating
your journey in data visualization with Python.
Chapter 6
Making Things
Interactive with Bokeh
Learning Objectives
Use Bokeh to create insightful web-

based visualizations
Explain the difference between the

two interfaces for plotting
Identify when to use the Bokeh

Server
Create interactive visualizations
In this chapter, we will design interactive plots using the Bokeh

library.
Introduction
Bokeh has been around since 2013, with version 1.0.4 being
released in 2018. It targets modern web browsers to present
interactive visualizations to users rather than static images. The
following are some of the features of Bokeh:
Simple visualizations: Through its
different interfaces, it targets users
of many skill levels, thus providing
an API for quick and simple
visualizations but also more
complex and extremely
customizable ones.
Excellent animated
visualizations: It provides high
performance and therefore can
work on large or even streaming
datasets, which makes it the go-to
choice for animated visualizations
and data analysis.
Inter-visualization interactivity:
This is a web-based approach,
where it's easy to combine several
plots and create unique and
impactful dashboards with
visualizations that can be
interconnected to create inter-
visualization interactivity.
Supports multiple languages:

Other than Matplotlib and
geoplotlib, Bokeh has libraries not
only for Python but also JavaScript
itself, and several other popular
languages.
Mulitiple ways to perform a task:
The previously mentioned
interactivity can be added in several
ways. The simplest inbuilt way is
the possibility to zoom and pan
around in your visualization, and
this already gives the user better
control of what they want to see.
Other than that, we can also
empower the users to filter and
transform the data.
Beautiful chart styling: The tech

stack is based on Tornado in the
backend and powered by D3 in the
frontend, which explains the
beautiful default styling of the
charts.
Since we are using Jupyter Notebook throughout this book, it's

worth mentioning that Bokeh, including its interactivity, is
natively supported in Notebook.
CONCEPTS OF BOKEH
The basic concept of Bokeh, in some ways, is comparable to
that of Matplotlib. In Bokeh, we have a figure as our root
element, which has sub-elements such as a title, an axis, and
glyphs. Glyphs have to be added to a figure, which can take on
different shapes such as circles, bars, and triangles to represent
a figure. The following hierarchy shows the different concepts
of Bokeh:
Figure 6.1: Concepts of Bokeh
INTERFACES IN BOKEH
The interface-based approach provides different levels of
complexity for users that either simply want to create some
basic plots with very few customizable parameters or the users
who want full control over their visualizations and want to
customize every single element of their plots. This layered
approach is divided into two levels:
Plotting: This layer is
customizable.
Models interface: This layer is

complex and provides an open
approach to designing charts.
Note
The models interfaces are the basic

building blocks for all plots.
The following are the two levels of the layered approach in

interfaces:
bokeh.plotting
This mid-level interface has a

somewhat comparable API to
Matplotlib. The workflow is to
create a figure and then enrich this
figure with different glyphs that
render data points in the figure.
Comparable to Matplotlib, the
composition of sub-elements such
as axes, grids, and the Inspector
(they provide basic ways of
exploring your data through
zooming, panning, and hovering) is
done without additional
configuration.
The important thing to note here is
that, even though its setup is done
automatically, we are able to
configure those sub-elements.
When using this interface, the
creation of the scene graph, which
is used by BokehJS, is handled
automatically, too.
bokeh.models
This low-level interface is

composed of two libraries: the
JavaScript library called BokehJS,
which gets used for displaying the
charts in the browser and the
Python library, which provides the
developer interface. Internally, the
definition created in Python creates
JSON objects that hold the
declaration for the JavaScript
representation in the browser.
The models interface exposes

complete control over how Bokeh
plots and widgets (that are elements
that enable users to interact with the
data displayed) are assembled and
configured. This means that it is up
to the developer to ensure the
correctness of the scene graph
(which is a collection of objects
describing the visualization).
OUTPUT
Outputting Bokeh charts is straightforward. There are three
ways this can be done, depending on your needs:
The .show() method: The basic

option is to simply display the plot
in an HTML page. This is done
with the .show() method.
Inline .show() method: When

working with Jupyter Notebook, the
.show() method will allow you
to display the chart inside your
notebook (when using inline
plotting).
The .output_file() method:

You're also able to directly save the
visualization to a file without any
overhead using the
.output_file() method. This
will create a new file at the given
path with a given name.
The most powerful way of providing your visualization is

through the use of the Bokeh Server.
BOKEH SERVER
As we mentioned before, Bokeh creates scene graph JSON
objects that will be interpreted by the BokehJS library to create
the visualization output. This process allows you to have a
unified format for other languages to create the same Bokeh
plots and visualizations, independent of the language used.
If we think one step further, what if we could also keep the

visualizations in sync with one another? This would enable us
to create way more complex visualizations and leverage the
tooling that Python provides. We could not only filter data but
also do calculations and operations on the server side, which
would update the visualizations in real time.
In addition to that, since we'd have an entry point for data, we

can create visualizations that get fed by streams instead of
static datasets. This design enables us to have way more
complex systems with even greater capabilities.
Looking at the scheme of this architecture, we can see that the

documents are provided on the server side, then moved over to
the browser client, which then inserts it into the BokehJS
library. This insertion will trigger the interpretation by
BokehJS, which will then create the visualization itself. The
following diagram describes the working of the Bokeh Server:
Figure 6.2: Workings of the Bokeh Server
PRESENTATION
In Bokeh, presentations help make the visualization more
interactive by using different features such as interactions,
styling, tools, and layouts.
Interactions
Probably the most interesting feature of Bokeh is its

interactions. There are basically two types of interactions:
passive and active.
Passive interactions are actions that the users can take that
neither change the data nor the displayed data. In Bokeh, this is
called the Inspector. As we mentioned before, the inspector
contains attributes such as zooming, panning, and hovering
over data. This tooling allows the user to inspect its data
further and might get better insights by only looking at a
zoomed-in subset of the visualized data points.
Active interactions are actions that directly change the

displayed data. This incorporates actions such as selecting
subsets of data or filtering the dataset based on parameters.
Widgets are the most prominent of active interactions, since
they allow users to simply manipulate the displayed data with
handlers. Widgets can be tools such as buttons, sliders, and
checkboxes. Referencing back to the subsection about the
output styles, those widgets can be used in both—the so-called
standalone applications and the Bokeh Server. This will help us
consolidate the recently learned theoretical concepts and make
things clearer. Some of the interactions in Bokeh are tab panes,
dropdowns, multi-selects, radio groups, text inputs, check
button groups, data tables, and sliders.
INTEGRATING
Embedding Bokeh visualizations can take two forms, as
follows:
HTML document: These are the standalone HTML

documents. These documents are very much self-contained.
Bokeh applications: They are backed by a Bokeh Server,

which means that they provide the possibility to connect, for
example: Python tooling for more advanced visualizations.
Bokeh is a little complex compared to Matplotlib with Seaborn

and has its drawbacks like every other library, but once you
have the basic workflow down, you're able to leverage the
benefits that come with Bokeh, which are the ways you can
extend the visual representation by simply adding interactivity
features and giving power to the user.
Note
One interesting feature is the to_bokeh method, which
allows you to plot Matplotlib figures with Bokeh without
configuration overhead. Further information about this method
is available at the following link:
https://bokeh.pydata.org/en/0.12.3/docs/user_guide/compat.ht
ml.
In the following exercises and activities, we'll consolidate the

theoretical knowledge and build several simple visualizations
to understand Bokeh and its two interfaces. After we've
covered the basic usage, we will compare the plotting and the
models interface to see the difference in using them and work
with widgets that add interactivity to the visualizations.
Note
All the exercises and activities in this chapter are developed
using Jupyter Notebook and Jupyter Lab. The files can be
downloaded from the following link: https://bit.ly/2T3Afn1.
EXERCISE 9: PLOTTING
WITH BOKEH
In this exercise, we want to use the higher-level interface that
is focused around providing a simple interface for quick
visualization creation. Refer to the introduction to check back
with the different interfaces of Bokeh. In this exercise, we will
be using the world_population dataset. This dataset
shows the population of different countries over the years. We
will use the plotting interface to get some insights into the
population densities of Germany and Switzerland:
1. Open the
exercise09_solution.ipyn
b Jupyter Notebook from the
this exercise. To do that, you need
to navigate to the path of this file in
the command-line terminal and
type in jupyter-lab.
2. As we described previously, we will

be using the plotting interface for
this exercise. The only elements we
have to import from plotting are the
figure (which will initialize a plot)
and the show method (which
displays the plot):
# importing the
import pandas as pd
from bokeh.plotting
import figure, show
3. One thing to note is that if we want

to display our plots inside a Jupyter
Notebook, we also have to import
and call the output_notebook
method from the io interface of
Bokeh:
# make bokeh display

figures inside the
notebook
from bokeh.io import

output_notebook
output_notebook()
4. Use pandas to load our
world_population dataset:

with pandas
dataset =
index_col=0)
5. A quick test by calling head on our

DataFrame shows us that our data
has been successfully loaded:
# looking at the
dataset
dataset.head()

Figure 6.3: Loading the top five rows of
the world_population dataset using he
head method
6. To populate our x and y axis, we

need to do some data extraction.
The x-axis will hold all the years
that are present in our columns. The
y-axis will hold the population
density values of the countries.
We'll start with Germany:
# preparing our data

for Germany
years = [year for year

in dataset.columns if
not year[0].isalpha()]
de_vals =
[dataset.loc[['Germany
']][year] for year in
years]
7. After extracting the wanted data,

we can create a new plot by calling
the Bokeh figure method. By
providing parameters such as
title, x_axis_label, and
y_axis_label, we can define
the descriptions displayed on our
plot. Once our plot is created, we
can add glyphs to it. In our
example, we will use a simple
line.By providing the legend
parameter next to the x and y
values, we get an informative
legend in our visualization:
# plotting the
population density
change in Germany in
the given years
plot =
figure(title='Populati
on Density of
Germany',
x_axis_label='Year',
y_axis_label='Populati
on Density')
plot.line(years,
de_vals, line_width=2,
legend='Germany')
show(plot)

Figure 6.4: Creating a line plot from
population density data of Germany
8. We now want to add another

country. In this case, we will use
Switzerland. We will use the
same technique that we used with
Germany to extract the data for
Switzerland:
# preparing the data

for the second country
ch_vals =
[dataset.loc[['Switzer
land']][year] for year
in years]
9. We can simply add several layers of

glyphs on to our figure plot. We can
also stack different glyphs on top of
one another, thus giving specific
data-improved visuals. In this case,
we want to add an orange line to
our plot that displays the data from
Switzerland. In addition to
that, we also want to have circles
for each entry in the data, giving us
a better idea about where the actual
data points reside. By using the
same legend name, Bokeh creates a
combined entry in the legend:
# plotting the data
for Germany and
Switzerland in one
visualization,
# adding circles for

each data point for
Switzerland
plot =
on Density of Germany
and Switzerland',
on Density')
plot.line(years,
de_vals, line_width=2,
legend='Germany')
plot.line(years,
ch_vals, line_width=2,
color='orange',
legend='Switzerland')
plot.circle(years,
ch_vals, size=4,
line_color='orange',
fill_color='white',
legend='Switzerland')
show(plot)
Figure 6.5: Adding Switzerland to the
plot
10. When looking at a larger amount of

data for different countries, it
makes sense to have a plot for each
of them separately. This can be
achieved by using one of the layout
interfaces. In this case, we are using
the gridplot:
# plotting the Germany

and Switzerland plot
in two different
visualizations
# that are
interconnected in
terms of view port
from bokeh.layouts
import gridplot
plot_de =
on Density of
Germany',
on Density',
plot_height=300)
plot_ch =
on Density of
Switzerland',
on Density',
plot_height=300,
x_range=plot_de.x_rang
e,
y_range=plot_de.y_rang
e)
plot_de.line(years,
de_vals, line_width=2)
plot_ch.line(years,
ch_vals, line_width=2)
plot =
gridplot([[plot_de,
plot_ch]])
show(plot)

Figure 6.6: Using a gridplot to display
the country plots next to each other
11. As the name suggests, we can

arrange the plots in a grid. This also
means that we can quickly get a
vertical display when changing the
two-dimensional list that was
passed to the gridplot method:
# plotting the above

declared figures in a
vertical manner
plot_v =
gridplot([[plot_de],
[plot_ch]])
show(plot_v)
The following screenshot shows the
Figure 6.7: Using the gridplot method to arrange the
visualizations vertically
Congratulations! We just covered the very basics of Bokeh.

Using the plotting interface makes it really easy to get
some quick visualizations in place. This really helps you
understand the data you're working with.
This simplicity, however, is achieved by abstracting away

complexity. We lose a lot of control by using the plotting
interface. In the next exercise, we'll compare the plotting
and models interfaces to show you how much abstraction is
added to plotting.
EXERCISE 10:
COMPARING THE
PLOTTING AND MODELS
INTERFACES
In this exercise, we want to compare the two interfaces:
plotting and models. We will compare them by creating a
basic plot with the high-level plotting interface and then
recreate this plot by using the lower-level models interface.
This will show us the differences between these two interfaces
and give us a good direction for the later exercises to
understand how to use the models interface:

b from Lesson06 to implement
this exercise. To do that, once
again, you need to navigate to the
path of this file. In the command-
line terminal, type jupyter-lab.
2. As we described previously, we will
be using the plotting interface
for this exercise. The only elements
we have to import from plotting are
the figure (which will initialize a
plot) and the show method (which
displays the plot):
# importing the
import numpy as np
import pandas as pd
3. One thing to note is that if we want

to display our plots inside a Jupyter
Notebook, we also have to import
and call the output_notebook
Bokeh:

figures inside the
notebook

output_notebook
output_notebook()
4. As we have done several times

before, we will use pandas to load
our world_population
dataset:

with pandas
dataset =
index_col=0)

# looking at the
dataset
dataset.head()

the world_population dataset using the
head method
6. In this part of the exercise, we will

work with the previously seen
plotting interface. As we saw
before, we basically only need to
import the figure to create, and
show to display our plot:
# importing the
plotting dependencies
from bokeh.plotting
import figure, show
7. Our data stays the same for both

plots, since we only want to change
the way we create our visualization.
We need the years present in the
dataset as a list, the mean
population density for the whole
dataset for each year, and the mean
population density per year for
Japan:
# preparing our data

of the mean values per
year and Japan
years = [year for year

in dataset.columns if
not year[0].isalpha()]
mean_pop_vals =
[np.mean(dataset[year]
) for year in years]
jp_vals =
[dataset.loc[['Japan']
][year] for year in
years]
8. When using the plotting

interface, we can create a plot
element that gets all the attributes
about the plot itself, such as the title
and axis labels. We can then use the
plot element and "apply" our
glyphs elements to it. In this case,
we will plot the global mean with a
line and the mean of Japan with
crosses:
# plotting the global

population density
change and the one for
Japan
plot =
figure(title='Global
Mean Population
Density compared to
Japan',
on Density')
plot.line(years,
mean_pop_vals,
line_width=2,
legend='Global Mean')
plot.cross(years,
jp_vals,
legend='Japan',
line_color='red')
show(plot)

Figure 6.9: Line plots comparing the global mean
population density with that of Japan
As we can see in the preceding diagram, we have many

elements already in place. This means that we already have the
right x-axis labels, the matching range for the y-axis, and our
legend is nicely placed in the upper-right corner without much
configuration.
Using the models Interface
1. The models interface is of a much

lower level compared to other
interfaces. We can already see this
when looking at the list of imports
we need for a comparable plot.
Looking through the list, we can
see some familiar names such as
Plot, Axis, Line, and Cross:
# importing the models

dependencies

show
from
bokeh.models.grids
import Grid
from
bokeh.models.plots
import Plot
from bokeh.models.axes
import LinearAxis
from
bokeh.models.ranges
import Range1d
from
bokeh.models.glyphs
import Line, Cross
from
bokeh.models.sources
import
ColumnDataSource
from
bokeh.models.tickers
import
SingleIntervalTicker,
YearsTicker
from
bokeh.models.renderers
import GlyphRenderer
from
bokeh.models.annotatio
ns import Title,
Legend, LegendItem
2. Before we build our plot, we have

to find out the min and max values
for the y-axis since we don't want to
have a too large or small range of
values. We therefore get all the
mean values for global and Japan
without any invalid values and then
get their smallest and largest values.
Those values are then passed to the
constructor of Range1d, which
will give us a range that can later be
used in the plot construction. For
the x-axis, we have our list of years
pre-defined:
# defining the range

for the x and y axis
extracted_mean_pop_val
s = [val for i, val in
enumerate(mean_pop_val
s) if i not in [0,
len(mean_pop_vals) -
1]]
extracted_jp_vals =
[jp_val['Japan'] for
i, jp_val in
enumerate(jp_vals) if
i not in [0,
len(jp_vals) - 1]]
min_pop_density =
min(extracted_mean_pop
_vals)
min_jp_densitiy =
min(extracted_jp_vals)
min_y =
int(min(min_pop_densit
y, min_jp_densitiy))
max_pop_density =
max(extracted_mean_pop
_vals)
max_jp_densitiy =
max(extracted_jp_vals)
max_y =
int(max(max_jp_densiti
y, max_pop_density))
xdr =
Range1d(int(years[0]),
int(years[-1]))
ydr = Range1d(min_y,
max_y)
3. Once we have the min and max

values for the y-axis, we can create
two Axis objects that will be used
to display the axis lines and the
label for the axis. Since we also
want ticks between the different
values, we have to pass in
a Ticker object that creates this
setup for us:
# creating the axis
axis_def =
dict(axis_line_color='
#222222',
axis_line_width=1,
major_tick_line_color=
'#222222',
major_label_text_color
='#222222',major_tick_
line_width=1)
x_axis =
LinearAxis(ticker =
SingleIntervalTicker(i
nterval=10),
axis_label = 'Year',
**axis_def)
y_axis =
LinearAxis(ticker =
SingleIntervalTicker(i
nterval=50),
axis_label =
'Population Density',
**axis_def)
4. Creating the title and plot itself is

straightforward. We can pass
a Title object to the title attribute
of the Plot object:
# creating the plot

object
title = Title(align =
'left', text = 'Global
Mean Population
Density compared to
Japan')
plot =
Plot(x_range=xdr,
y_range=ydr,
plot_width=650,
plot_height=600,
title=title)
5. If we try to display our plot now

using the show method, we will get
an error, since we have no renderers
defined at the moment. First, we
need to add elements to our plot:
# error will be thrown

because we are missing
renderers that are
created when adding
elements
show(plot)

Figure 6.10: Empty plot with title
6. When working with data, we

always need to insert our data into a
DataSource object. This can then be
used to map the data source to the
Glyph object that will be displayed
in the plot:
# creating the data

display
line_source =
ColumnDataSource(dict(
x=years,
y=mean_pop_vals))
line_glyph =
Line(x='x', y='y',
line_color='#2678b2',
line_width=2)
cross_source =
ColumnDataSource(dict(
x=years, y=jp_vals))
cross_glyph =
Cross(x='x', y='y',
line_color='#fc1d26')
7. When adding objects to the plot,

you have to use the right add
method. For layout elements such
as the Axis objects, we have to use
the add_layout method.
Glyphs, which display our data,
have to be added with the
add_glyph method:
# assembling the plot
plot.add_layout(x_axis
, 'below')
plot.add_layout(y_axis
, 'left')
line_renderer =
plot.add_glyph(line_so
urce, line_glyph)
cross_renderer =
plot.add_glyph(cross_s
ource, cross_glyph)
8. If we now try to show our plot, we

can finally see that our lines are in
place:
show(plot)

Figure 6.11: Models interface-based plot
displaying the lines and axes
9. A few elements are still missing.

One of them is the legend in the
upper-right corner. To add a legend
to our plot, we again have to use an
object. Each LegendItem object
will be displayed in one line in the
legend:
# creating the legend
legend_items=
[LegendItem(label='Glo
bal Mean', renderers=
[line_renderer]),
LegendItem(label='Japa
n', renderers=
[cross_renderer])]
legend =
Legend(items=legend_it
ems,
location='top_right')
10. Creating the grid is straightforward:

we simply have to instantiate
two Gridobjects for the x and y
axes. These grids will get the
tickers of the previously created x
and y axes:
# creating the grid
x_grid =
Grid(dimension=0,
ticker=x_axis.ticker)
y_grid =
Grid(dimension=1,
ticker=y_axis.ticker)
11. To add the last final touches, we,

again, use
the add_layout method to add
the grid and the legend to our plot.
After this, we can finally display
our complete plot, which will look
like the one we created in the first
task, with only four lines of code:
# adding the legend

and grids to the plot
plot.add_layout(legend
)
plot.add_layout(x_grid
)
plot.add_layout(y_grid
)
show(plot)

Figure 6.12: Full recreation of the visualization done
with the plotting interface
Congratulations! As you can see, the models interface should

not be used for simple plots. It's meant to provide the full
power of Bokeh to experienced users that have specific
requirements that need more than the plotting interface.
Having seen the models interface before will come in handy
in our next topic, which is about widgets.
Adding Widgets
One of the most powerful features of Bokeh is its ability to use
widgets to interactively change the data that's displayed in the
visualization. To understand the importance of interactivity in
your visualizations, imagine seeing a static visualization about
stock prices that only show data for the last year. If this is what
you specifically searched for, it's suitable enough, but if you're
interested to see the current year or even visually compare it to
recent years, those plots won't work and will add additional
work, since you have to create them for every year. Comparing
this to a simple plot that lets the user select the wanted date
range, we can already see the advantages. There are endless
options to combine widgets and tell your story. You can guide
the user by restricting values and only displaying what you
want them to see. Developing a story behind your visualization
is very important, and doing this is much easier if the user has
ways of interacting with the data.
Bokeh widgets work best when used in combination with the

Bokeh server. However, using the Bokeh server approach is
beyond the content of this book, since we would need to work
with simple Python files and can't leverage the power of
Python notebook. Instead, we will use a hybrid approach that
only works with the older Jupyter Notebook.
EXERCISE 11: BASIC

INTERACTIVITY WIDGETS
This first exercise of the Adding Widgets topic will give you a
gentle introduction into the different widgets and the general
concept of how to use them in combination with your
visualizations. We will look at basic widgets and build a simple
plot that will display the first 25 data points for the selected
stock. The displayed stock can be changed with a drop-down
menu.
The dataset of this exercise is a stock_prices dataset. This

means that we will be looking at data over a range of time. As
this is a large and variable dataset, it will be easier to show and
explain different widgets such as slider and dropdown on it.
The dataset is available in the data folder of the GitHub
repository; here is the link to it: https://bit.ly/2UaLtSV.
We will look at the different available widgets and how to use

them before going in and building a basic plot with one of
them. There are a few different options regarding how to
trigger updates, which are also explained in the following
steps. The widgets that will be covered in this exercise are
explained in the following table:
Figure 6.13: Some of the basic widgets with
examples
1. Open the
this exercise. Since we need to user
Jupyter Notebook in this example,
we will type in the following at the
command line: jupyter
notebook.
2. A new browser window will open

that lists all the files in the current
directory. Click on
exercise11_solution.ipynb
, which will open it in a new tab.
3. This exercise will first introduce

you to the basic widgets before
showing you how to create a basic
visualization with them. Therefore,
we will add more imports in
suitable places throughout the code.
For importing our dataset, we need
pandas:
# importing the
import pandas as pd
4. Again, we want to display our plots
inside a Jupyter Notebook, so we
have to import and call the
output_notebook method from
the io interface of Bokeh:

figures inside the
notebook

output_notebook
output_notebook()
5. After downloading the dataset and

moving it into the data folder of this
chapter, we can import our
stock_prices.csv data:
# loading the Dataset

with geoplotlib
dataset =
pd.read_csv('./data/st
ock_prices.csv')

# looking at the
dataset
dataset.head()

the stock_prices dataset using the head
method
7. Since the date column has no

information about the hour, minute,
and second, we want to avoid
displaying them in the visualization
later on and simply display the year,
month, and day. Therefore, we'll
create a new column that holds the
formatted short version of the date
value. Note that the execution of
the cell will take a moment, since
it's a fairly large dataset. Please be
patient:
# mapping the date of

each row to only the
year-month-day format
from datetime import

datetime
def
shorten_time_stamp(tim
estamp):
shortened =
timestamp[0]
if len(shortened) >
10:
parsed_date=datetime.s
trptime(shortened,
'%Y-%m-%d %H:%M:%S')
shortened=datetime.str
ftime(parsed_date,
'%Y-%m-%d')
return shortened
dataset['short_date']
= dataset.apply(lambda
x:
shorten_time_stamp(x),
axis=1)
8. Taking another look at our updated

dataset, we can see a new column
called short_date that holds the
date without the hour, minute, and
second information:
# looking at the
dataset with shortened
date
dataset.head()

Figure 6.15: Dataset with the added short_date

column
Looking at Basic Widgets
1. In this first task, the interactive

widgets are added with the interact
element of IPython. We have to
specifically import them:
# importing the
widgets
from ipywidgets import

interact,
interact_manual
2. We'll be using the "syntactic sugar"

way of adding a decorator to a
method, that is, by using
annotations. This will give us an
interactive element that will be
displayed below the executable cell.
In this example, we'll simply print
out the result of the interactive
element:
# creating a checkbox
@interact(Value=False)
def
checkbox(Value=False):
print(Value)

Figure 6.16: Interactive checkbox that
will switch from False to True if checked
Note
@interact() is called a
decorator, which wraps the
annotated method into the interact
component. This allows us to
display and react to the change of
the drop-down menu. The method
will be executed every time the
value of the dropdown changes.
3. Once we have the first element in

place, all the other elements are
created the exact same way, just
altering the data type of the
arguments in the decorator.
4. Use the following code for the

dropdown:
# creating a dropdown
options=['Option1',
'Option2', 'Option3',
'Option4']
@interact(Value=option
s)
def
slider(Value=options[0
]):
print(Value)

Figure 6.17: Interactive dropdown
Use the following code for the

input text:
# creating an input
text
@interact(Value='Input
Text')
def slider(Value):
print(Value)

Figure 6.18: Interactive text input
Use the following code to apply

multiple widgets:
# multiple widgets
with default layout
options=['Option1',
'Option2', 'Option3',
'Option4']
@interact(Select=optio
ns, Display=False)
def uif(Select,
Display):
print(Select, Display)

Figure 6.19: Two widgets are displayed

vertically by default
Use the following code to apply an

int slider:
# creating an int
slider with dynamic
updates
@interact(Value=(0,
100))
def slider(Value=0):
print(Value)

Figure 6.20: Interactive int slider
Use the following code to apply an

int slider that triggers upon
releasing the mouse:
# creating an int
slider that only
triggers on mouse
release

IntSlider
slider=IntSlider(min=0
, max=100,
continuous_update=Fals
e)
@interact(Value=slider
)
def slider(Value=0.0):
print(Value)

Figure 6.21: Interactive int slider that

only triggers upon mouse release
Note:
Although the output of Figures 6.20

and 6.21 looks the same, in Figure
6.21, the slider triggers only upon
mouse release.
5. If we don't want to update our plot

every time we change our widget,
we can also use the
interact_manual decorator,
which adds an execution button to
the output:
# creating a float
slider 0.5 steps with
manual update trigger
@interact_manual(Value
=(0.0, 100.0, 0.5))
def slider(Value=0.0):
print(Value)

Figure 6.22: Interactive int slider with a manual

update trigger
Note
Compared to the previous cells, this one contains the
interact_manual decorator instead of interact. This will
add an execution button, which will trigger the update of the
value instead of triggering with every change. This can be
really useful when working with larger datasets, where the
recalculation time would be large. Because of this, you don't
want to trigger the execution for every small step but only once
you have selected the right value.
Creating a Basic Plot and Adding a Widget
In this task, we will create a basic visualization with the stock

price dataset. This will be your first interactive visualization in
which you can dynamically change the stock that is displayed
in the graph. We will get used to one of the aforementioned
interactive widgets: the drop-down menu. It will be the main
point of interaction for our visualization:
1. To be able to create a plot, we first

need to import the already-familiar
figure and show methods from the
plotting interface. Since we also
want to have a panel with two tabs
displaying different plot styles, we
also need the Panel and Tabs
classes from the models interface:
# importing the
from
bokeh.models.widgets
import Panel, Tabs
from bokeh.plotting
import figure, show
2. To structure our notebook better,

we want to write an adaptable
method that gets a subsection of
stock data as an argument and
builds a two-tab Pane object that
lets us switch between the two
views in our visualization. The first
tab will contain a line plot of the
given data, while the second one
will contain a circle-based
representation of the same data. A
legend will display the name of the
currently viewed stock:
# method to build the

tab-based plot
def get_plot(stock):
stock_name=stock['symb
ol'].unique()[0]
line_plot=figure(title
='Stock prices',
x_axis_label='Date',
x_range=stock['short_d
ate'],
y_axis_label='Price in
$USD')
line_plot.line(stock['
short_date'],
stock['high'],
legend=stock_name)
line_plot.xaxis.major_
label_orientation = 1
circle_plot=figure(tit
le='Stock prices',
x_axis_label='Date',
x_range=stock['short_d
ate'],
y_axis_label='Price in
$USD')
circle_plot.circle(sto
ck['short_date'],
stock['high'],
legend=stock_name)
circle_plot.xaxis.majo
r_label_orientation =
1
line_tab=Panel(child=l
ine_plot,
title='Line')
circle_tab=Panel(child
=circle_plot,
title='Circles')
tabs = Tabs(tabs=[
line_tab, circle_tab
])
return tabs
3. Before we can build our interaction,

we have to get a list of all the stock
names that are present in the
dataset. Once we have done that,
we can then use this list as an input
for the interact element. With each
interaction of the dropdown, our
displayed data will then be updated.
To keep it simple, we only want to
display the first 25 entries of each
stock in this task. By default, the
stock of Apple should be displayed.
Its symbol in the dataset is AAPL:
# extracing all the

stock names
stock_names=dataset['s
ymbol'].unique()
4. We can now add the drop-down

widget in the decorator and call the
method that returns our
visualization in the show method
with the selected stock. This will
give us a visualization that is
displayed in a pane with two tabs.
The first tab will display an
interpolated line and the second tab
will display the values as circles:
# creating the
dropdown interaction
and building the plot
# based on selection
@interact(Stock=stock_
names)
def
get_stock_for(Stock='A
APL'):
stock =
dataset[dataset['symbo
l'] == Stock][:25]
show(get_plot(stock))

Figure 6.23: Line tab with the data of AAPL displayed
The following screenshot shows the output of the code in step

16:
Figure 6.24: Circle tab with the data of AAPL
displayed
Note
We can already see that each date is displayed on the x-axis. If
we want to display a bigger time range, we have to customize
the ticks on our x-axis. This can be done using ticker objects.
Congratulations! We just covered the very basics of widgets

and how to use them in Jupyter Notebook.
Note
If you want to learn more about using widgets and which
widgets can be used in Jupyter, you can refer to these links:
https://bit.ly/2Sx9txZ and https://bit.ly/2T4FcM1.

PLOTS WITH WIDGETS
In this activity, you will combine what you have already
learned about Bokeh. You will also need the skills you have
acquired while working with pandas for additional DataFrame
handling. We will create an interactive visualization that lets us
explore the end results of the Olympics 2016 in Rio.
Our visualization will display each country that participated in

a coordinate system where the x-axis represents the number of
won medals and the y-axis represents the number of athletes.
Using interactive widgets, we will be able to filter down the
displayed countries in both the maximum amount of won
medals and the maximum amount of athletes axes.
There are many options when it comes to choosing which

interactivity to use. We will focus on having only two widgets
to make it easier for you to understand the concepts. In the end,
we will have a visualization that allows us to filter countries for
the amount of medals and athletes they placed in the Olympics
and upon hovering over the single data points, receive more
information about each country:

this activity.
2. Don't forget to enable to notebook

output using the bokeh.io
interface. Import pandas and load
the dataset. Make sure that the
dataset is loaded by displaying the
first five elements of the dataset.
3. Import figure and show from

Bokeh and interact and
widgets from ipywidgets to
get started.
4. When starting to create our

visualization, we have to import the
tools we will be using. Import
figure and show from Bokeh
and the interact and widgets
interfaces from ipywidgets.
5. After extracting the necessary data,

we will set up the interaction
elements. Scroll down until you
reach the cell that says getting
the max amount of medals
and athletes of all
countries. Extract those two
numbers from the dataset.
6. After extracting the maximum

numbers of medals and athletes,
create widgets for IntSlider for
the maximum number of athletes
(orientation vertical) and
IntSlider for the maximum
number of medals (orientation
horizontal).
7. The last step of preparation before

implementing our plotting is to set
up the @interact method, which
will display the whole visualization.
The only code we will write here is
to show the return value of the
get_plot method that gets all the
interaction element values as
parameters.
8. After implementing the decorated

method, we can now move up in
our notebook and work on the
get_plot method.
9. First, we want to filter down our

countries dataset that contains all
the countries that placed athletes in
the olympic games. We need to
check whether they have less or
equal number of medals and
athletes than our max values passed
as arguments.
10. Once we have a filtered down our

dataset, we can create our
DataSource. This DataSource will
be used both for the tooltips and the
printing of the circle glyphs.
11. After that, we will create a new plot

using the figure method that has the
following attributes: title as Rio
Olympics 2016 - Medal
comparison, x_axis_label as
Number of Medals,
y_axis_label as Num of
Athletes.
12. The last step is to execute every cell

starting from the get_plot cell to
the bottom, again, making sure that
all implementations are captured.
13. When executing the cell that

contains the @interact
decorator, you will see the scatter
plot that displays a circle for every
country displaying additional
information such as the short code
of the country, the amount of
athletes, and the number of gold,
silver, and bronze medals.
Note:

found on page 277.
Summary
In this chapter, we have looked at another option for creating
visualizations with a whole new focus: web-based Bokeh plots.
We also discovered ways in which we can make our
visualizations more interactive and really give the user the
chance to explore data in a whole different way. As we
mentioned in the first part of this chapter, Bokeh is a
comparably new tool that empowers developers to use their
favorite language to create easily portable visualizations for the
web. After working with Matplotlib, Seaborn, geoplotlib, and
Bokeh, we can see some common interfaces and similar ways
to work with those libraries. After understanding the tools that
are covered in this book, it will be simple to understand new
plotting tools.
In the next and final chapter, we will introduce a new, not-yet-

covered dataset, with which you will create a visualization.
This last chapter will allow you to consolidate the concepts and
tools that you have learned about in this book and further
enhance your skills.
Chapter 7
Combining What We
Have Learned
Learning Objectives
Apply your skills in Matplotlib and

Seaborn
Create a time series with Bokeh
Analyze geospatial data with

geoplotlib
In this chapter, we will apply all the concepts that we have

learned in all the previous chapters. We will use three new
datasets in combination with practical activities for Matplotlib,
Seaborn, geoplotlib, and Bokeh. We will conclude this chapter
with a summary that recaps what we've learned throughout the
book.
Introduction
To consolidate what we have learned, we will provide you with
three sophisticated activities. Each activity uses one of the
libraries that we have covered in this book. Every activity has a
bigger dataset than we have used before in this book, which
will prepare you for larger datasets.
Note
All activities will be developed in the Jupyter Notebook or
Jupyter Lab. Please download the GitHub repository with all
the prepared templates from https://bit.ly/2SswjqE.
ACTIVITY 30:
IMPLEMENTING
MATPLOTLIB AND
SEABORN ON NEW YORK
CITY DATABASE
In this activity, we will visualize data about New York City
(NYC) and compare it to the state of New York and the United
States (US). The American Community Survey (ACS) Public-
Use Microdata Samples (PUMS) dataset (one-year estimate
from 2017) from https://www.census.gov/programs-
surveys/acs/technical-
documentation/pums/documentation.2017.html is used. For
this activity, you can either use Matplotlib, Seaborn, or a
combination of both.
In this activity, the datasets "New York Population Records"

(./data/pny.csv) and "New York Housing Unit Records"
(./data/hny.csv) are used. The first dataset contains
information about the New York population, and the second
dataset contains information about housing units. The dataset
contains data for about 1% of the population and housing units.
Due to the extensive amount of data, we do not provide the
datasets for the whole of the US; instead, we will provide the
required information related to the US if necessary. The
PUMS_Data_Dictionary_2017.pdf PDF gives an
overview and description of all variables. A further description
of the codes can be found in
ACSPUMS2017CodeLists.xls:

this activity.
2. Use pandas to read both .csv files

located in the subdirectory data.
3. Use the given PUMA (public use

microdata area code based on the
2010 Census definition, which are
areas with populations of 100k or
more) ranges to further divide the
dataset into NYC districts (Bronx,
Manhatten, Staten Island,
Brooklyn, and Queens):
# PUMA ranges
bronx = [3701, 3710]
manhatten = [3801,
3810]
staten_island = [3901,
3903]
brooklyn = [4001,
4018]
queens = [4101, 4114]
nyc = [bronx[0],
queens[1]]
4. In the dataset, each sample has a

certain weight that reflects the
weight for the total dataset.
Therefore, we cannot simply
calculate the median. Use the given
weighted_median function in
the following code to compute the
median:
# Function for a
'weighted' median
def
weighted_frequency(val
ues, weights):
weighted_values = []
for value, weight in

zip(values, weights):
weighted_values.ex
tend(np.repeat(value,
weight))
return
weighted_values
def
weighted_median(values
, weights):
return
np.median(weighted_fre
quency(values,
weights))
5. In this subtask, we will create a plot

containing multiple subplots that
visualize information with regard to
NYC wages. Visualize the median
household income for the US, New
York, New York City, and its
districts. Visualize the average
wage by gender for the given
occupation categories for the
population of NYC:
occ_categories =
['Management,\nBusines
s,\nScience,\nand
Arts\nOccupations',
'Service\nOccupations'
,
'Sales
and\nOffice\nOccupatio
ns', 'Natural
Resources,\nConstructi
on,\nand
Maintenance\nOccupatio
ns',
'Production,\nTranspor
tation,\nand Material
Moving\nOccupations']
occ_ranges =
{'Management,
Business, Science, and
Arts Occupations':
[10, 3540], 'Service
Occupations': [3600,
4650],
'Sales and Office

5940], 'Natural
Resources,
Construction, and
Maintenance
7630],
'Production,
Transportation, and
Material Moving
9750]}
Visualize the wage distribution for

New York and NYC. Use the
following yearly wage intervals:
10k steps between 0 and 100k, 50k
steps between 100k and 200k, and
>200k.

percentage for the given occupation
subcategories for the population of
NYC:
occ_subcategories =
{'Management,\nBusines
s,\nand Financial':
[10, 950],
'Computer,
Engineering,\nand
Science': [1000,
1965],
'Education,\nLegal,\nC
ommunity
Service,\nArts,\nand
Media': [2000, 2960],
'Healthcare\nPractitio
ners\nand\nTechnical':
[3000, 3540],
'Service': [3600,
4650],
'Sales\nand Related':
[4700, 4965],
'Office\nand
Administrative\nSuppor
t': [5000, 5940],
'': [6000, 6130],
'Construction\nand
Extraction': [6200,
6940],
'Installation,\nMainte
nance,\nand Repair':
[7000, 7630],
'Production': [7700,
8965],
'Transportation\nand
Material\nMoving':
[9000, 9750]}
7. Use a heatmap to show the

correlation between difficulties
(self-care difficulty, hearing
difficulty, vision, difficulty,
independent living difficulty,
ambulatory difficulty, veteran
service-connected disability, and
cognitive difficulty) and age groups
(<5, 5-11, 12-14, 15-17, 18-24, 25-
34, 35-44, 45-54, 55-64, 65-74, and
75+) in New York City.
Note:
The solution to this activity can be

found on page 282.
BOKEH
Stock price data is one of the most interesting types of data for
many people. When thinking about its nature, we can see that it
is highly dynamic and constantly changing. To understand it,
we need high levels of interactivity to not only look at the
stocks of interest, but also to compare different stocks, see their
traded volume, and the high/lows of the given dates and
whether it rose or sunk the day before that.
Taking into account all of the aforementioned features, we

need to use a highly customizable visualization tool. We also
need the possibility to add different widgets to enable
interactivity. In this activity, we will therefore use Bokeh to
create a candle-stick visualization with several interactivity
widgets to enable better exploration of our data.

BOKEH STOCK PRICES
This activity will combine most of what you have already
learned about Bokeh. You will also need the skills you have
acquired while working with pandas. We will create an
interactive visualization that displays a candlestick plot, which
is often used when handling stock price data. We will be able
to compare two stocks to each other by selecting them from
dropdowns. A RangeSlider will allow us to restrict the
displayed date range in the requested year 2016. Depending on
what graph we choose, we will either see the candlestick
visualization or a simple line plot displaying the volume of the
selected stock:

this activity.
2. Don't forget to enable notebook

output by using the bokeh.io
interface. Import pandas and load
the downloaded dataset. Make sure
that the dataset is loaded by
displaying the first five elements of
the dataset.
3. We need to create a column in our

DataFrame that holds the
information from the date column
without the hour, minute, and
second information. Make sure
your implementation works
correctly by displaying the first five
elements of the updated DataFrame.
4. There are many options when it

comes to choosing which
interactivity to use. Since the goal
of this activity is to be able to
compare two stocks with each other
in terms of traded volume and the
high/low and open/close prices over
a time range, we will need widgets
to select elements and a slider to
select a given range. Considering
that we have two options of display,
we also need a way to select either
one or the other.
5. Import figure and show from

Bokeh and interact and
widgets from ipywidgets to
get started.
6. When starting to create our

visualization, we have to import the
tools we will be using. Import
figure and show from Bokeh
and the interact and widgets
interface from ipywidgets.
7. Execute the cells from top to

bottom until you reach the cell that
has the comment #extracting
the necessary data. Start
your implementation there. Get the
unique stock names from the
dataset. Filter out the dates from
2016. Only get the unique dates
from 2016. Create a list that
contains the strings open-close
and volume, which will be used
for the radio buttons to switch
between the two plots.
8. After extracting the necessary data,

we will set up the interaction
elements. Create widgets for the
following: Dropdown for the first
stock name (default value will be
AAPL); Dropdown for the second
stock name that will be compared
to the first (default value will be
AON);
SelectionRangeSlider to
select the range of dates we want to
display in our plot (default values
displayed will be 0 to 25);
RadioButtons to choose between
the candlestick plot and the plot
that displays the traded volume (the
default value will be open-
close, which will display the
candlestick plot.)
9. The last step of preparation before

implementing our plot is to set up
the @interact method that
finally displays the whole
visualization.
10. The only code we will write here is

to show the return value of the
get_plot method that gets all the
interaction element values as
parameters.
11. After implementing the decorated

method, we can now move up in
our notebook and work on the
add_candle_plot method.
Make sure to use the example in the
Bokeh documentation as an outline.
You can find the same in this link:
https://bokeh.pydata.org/en/latest/d
ocs/gallery/candlestick.html.
12. The next step will be to move on

and implement the line plot in the
cell that contains the get_plot
method. Plot a line for the data
from stock_1 with a blue color.
Plot a line for the data from
stock_2 with an orange color.
13. Before finalizing this activity, we

want to add one more interactivity
feature: muting different elements
of our plot. This can be done by
clicking on one of the displayed
elements in the legend of the
visualization. Of course, we first
have to tell Bokeh what it should
do. Read up on this at
ocs/user_guide/interaction/legends.
html.
14. The last step is to execute every cell

starting from the
add_candle_plot cell to the
bottom, again making sure that all
implementations are captured.
15. When executing the cell that

contains the @interact
decorator, you will see the
candlestick plot for the two, by
default, selected, with the AAPL
and AON stocks displayed.
Note:

found on page 288.
GEOPLOTLIB
The dataset that's used in this activity is of Airbnb, which is
publicly available online. Accommodation listings have two
predominant feature: latitude and longitude. Those two features
allow us to create geo-spatial visualizations that gives us a
better understanding of attributes such as the distribution of
accommodations across each city.
In this activity, we will therefore use geoplotlib to create a

visualization that maps each accommodation to a dot on a map,
colored based on either the price or rating of that listing. The
two attributes can be switched by pressing the left and right
keys on the keyboard.

AIRBNB DATA WITH
GEOPLOTLIB
In this last activity for geoplotlib, we will use airbnb listing
data to determine the most expensive and best-rated regions of
accommodation in the New York area. We will write a custom
layer with which we can switch between the price and the
review score of each accommodation. In the end, we will be
able to see the hotspots for the most expensive and best-rated
accommodation across New York.
In theory, we should see a price increase the closer we get to

the center of Manhatten. It will be very interesting to see
whether the ratings for the given accommodations also increase
the closer we get to the center of Manhatten:

this activity.
2. First, make sure you import the

necessary dependencies.
3. Load the
airbnb_new_york.csv dataset
using pandas. If your system is a
little bit slower, just use the
airbnb_new_york_smaller.
csv dataset with less data points.
4. Get a feeling for the dataset and the

features it has by taking a look at it.
5. Since our dataset once again has

columns that are named
Latitude and Longitude
instead of lat and lon, rename
those columns to their short
versions, which we need for
geoplotlib.
6. In addition to that, we also want to

clean and map our two major
coumns: price and
review_scores_rating. Fill
the n/a values and create a new
column called dollar_price
that holds the price as a float.
7. Before we finally create the layer,

we also want to trim down the
number of columns our working
dataset has. Create a subsection of
the columns with id, latitude
(as lat), longitude (as lon),
price (in $), and
review_scores_rating.
8. Create a new
DataAccessObject with the
newly created subsection of the
dataset. Use it to plot out a dot map.
9. Create a new ValueLayer that

extends the geoplotlib
BaseLayer.
10. Given the data, we want to plot

each point on the map with a color
that is defined by the currently
selected attribute, either price or
rating.
11. To assign each point a different

color, we simply paint each point
separately. This is definitely not the
most efficient solution, but it will
do for now.
12. We will need the following instance

variables: self.data, which
holds the dataset;
self.display, which holds the
currently selected attribute name;
self.painter, which holds an
instance of the BatchPainter
class; self.view, which holds
the BoundingBox; self.cmap,
which holds a color map with the
jet color schema; and an alpha of
255 and 100 levels.
13. Implement the __init__,

invalidate, draw, and bbox
method for the ValueLayer.
14. Use the provided BoundingBox

that's focused on New York when
calling the ValueLayer.
Note:

found on page 296.
Summary
This chapter gave us a short overview and recap of everything
that was covered in this bookware on the basis of three
extensive practical activities. In Chapter 1, The Importance of
Data Visualization and Data Exploration, we started with a
Python library journey that we used as a guide throughout the
whole bookware. We first talked about the importance of data
and visualizing this data to get meaningful insights into it and
gave a quick recap on different statistics concepts. In several
activities, we learned how to import and handle datasets with
Numpy and pandas. In Chapter 2, All You Need to Know about
Plots, we discussed various visualizations plots/charts and
which visualizations are best to display certain information. We
mentioned the use case, design practices, and practical
examples for each plot type.
In Chapter 3, A Deep Dive into Matplotlib, we thoroughly

covered Matplotlib and started with the basic concepts. Next, a
deep insight into the numerous possibilities to enrich
visualizations with text was given. Emphasis was put on
explaining almost all plotting functions Matplotlib offers using
practical examples. Furthermore, we talked about different
ways to create layouts. The chapter was rounded off with how
to visualize images and write mathematical expressions. In
Chapter 4, Simplifying Visualizations Using Seaborn, Seaborn
was covered, which is built on top of Matplotlib and provides a
higher-level abstraction to make insightful visualizations. With
several examples, we showed you how Seaborn can simplify
the creation of visualizations. We also introduced further plots,
such as heatmaps, violin plots, and correlograms. Finally, we
used Squarify to create tree maps.
Visualizing geospatial data was covered in Chapter 5, Plotting

Geospatial Data, using geoplotlib. Understanding how
geoplotlib is structured internally explained why we had to
work with the pyglet library when adding interactivity to our
visualizations. In the book of this chapter, we worked with
different datasets and built both static and interactive
visualizations for geospatial data. In Chapter 6, Making Things
Interactive with Bokeh, we focused on working with Bokeh,
which targets modern web browsers to present interactive
visualizations. Starting with simple examples, we emphasized
the biggest advantage of Bokeh, namely interactive widgets.
We ended the book with this chapter, applying all the skills that
we've learned about by using three real-life datasets.
Appendix
About
This section is included to assist the students to perform the
activities in the course. It includes detailed steps that are to be
performed by the students to achieve the objectives of the
activities.
Chapter 1: The
Importance of Data
Visualization and Data
Exploration
ACTIVITY 1: USING
NUMPY TO COMPUTE
THE MEAN, MEDIAN,
VARIANCE, AND
STANDARD DEVIATION
FOR THE GIVEN
NUMBERS
Solution:
Let's use NumPy to calculate the mean, median, variance, and

standard deviation:
1. Import the necessary libraries:
# importing the
import numpy as np
2. Load the
dataset by using the genfromtxt
method of NumPy:
dataset =
v', delimiter=',')
3. First, we want to print a subset of

the first two rows of the dataset:
# looking at the first

two rows of the
dataset
dataset[0:2]

as follows:
Figure 1.22: First two rows of the

dataset
4. Once we know that our dataset has

been successfully loaded, we can
start solving our first task, which is
calculating the mean of the third
row.
The third row can be accessed by

indexing dataset[2]:
# calculate the mean

of the third row
np.mean(dataset[2])

as follows:
Figure 1.23: Mean of the third row
5. The last element of an ndarray can

be indexed the same way a regular
Python List can be accessed.
dataset[:, -1] will give us
the last column of every row:

of the last column
np.mean(dataset[:,-1])

as follows:
Figure 1.24: Mean of the last column
6. The double-indexing mechanism of

NumPy gives us a nice interface to
extract sub-selection. In this task,
we are asked to get a sub-matrix of
the first three elements of every
row of the first three columns:

of the intersection of
the first 3 rows and
first 3 columns
np.mean(dataset[0:3,
0:3])

as follows:
Figure 1.25: Mean of an intersection
7. Moving to the next set of tasks that

cover the usage of the median, we
will see that the API is consistent
between the different methods.:
# calculate the median

of the last row
np.median(dataset[-1])

as follows:
Figure 1.26: Median of the last row

8. Reverse indexing also works when
defining a range. So, if we want to
get the last three columns, we can
use dataset[:, -3:]:

of the last 3 columns
np.median(dataset[:,
-3:])

as follows:
Figure 1.27: Median of the last 3

columns
9. As we saw in the previous exercise,

we can aggregate the values along
an axis. If we want to calculate
the rows, we use axis=1:

of each row
np.median(dataset,
axis=1)

as follows:
Figure 1.28: Using axis to calculate the
median of each row
10. The last method we'll cover here is

variance. Again, NumPy provides
us with a consistent API, which
makes doing this easy. To calculate
the variance for each column, we
have to use axis 0:
# calculate the
variance of each
column
np.var(dataset,
axis=0)

as follows:
Figure 1.29: Variance across each

column
11. When only looking at a very small
subset of the matrix (2x2) elements,
we can apply what we learned in
the statistical overview to observe
that the value is way smaller than
the whole dataset:
Note
A small subset of a dataset does not

display the attributes of the whole.
# calculate the
variance of the
intersection of the
last 2 rows and first
2 columns
np.var(dataset[-2:,
:2])

as follows:
Figure 1.30: Variance of a small subset

of the dataset
12. The values of the variance might

seem a little bit strange at first.
You can always go back to the

Measures of Dispersion topic to
recap what you've learned so far.
Just remember, variance is not the
standard deviation:
# calculate the
standard deviation for
the dataset
np.std(dataset)

as follows:
Figure 1.31: Standard deviation of the complete

dataset

NumPy. In the following activities, this knowledge will be
further consolidated.
ITERATING
Solution:
Let's use the features of NumPy to index, slice, split, and

iterate ndarrays.
Indexing
# importing the
import numpy as np
2. Load the
look at the ndarray, like in the
previous activity:
dataset =
v', delimiter=',')
3. First, we want to use simple

indexing for the second row, as we
did in our first exercise. For a
clearer understanding, all the
elements are saved to a variable:
# indexing the second

row of the dataset
(second row)
second_row =
dataset[1]
np.mean(second_row)

as follows:
Figure 1.32: A screenshot of the mean of
the whole second row
4. Now, we need to reverse index the

last row and calculate the mean of
that row. Always remember that
providing a negative number as the
index value will index the list from
the end:
# indexing the last

element of the dataset
(last row)
last_row = dataset[-1]
np.mean(last_row)

as follows:
Figure 1.33: Calculating the mean of the

last row
5. Two-dimensional data can be

accessed the same as with a Python
List by using [0][0], where the
first pair of brackets accesses the
row and the second one accesses
the column.
However, we can also use comma-

separated notation, like [0, 0]:
# indexing the first

value of the second
row (1st row, 1st
value)
first_val_first_row =
dataset[0][0]
np.mean(first_val_firs
t_row)

as follows:
Figure 1.34: Calculating the mean of a

single value won't throw an error
6. The last value of the second last

row can easily be accessed with
reverse indexing. Remember that
-1 means the last element:
# indexing the last

value of the second to
last row (we want to
use the combined
access syntax here)
last_val_second_last_r
ow = dataset[-2, -1]
np.mean(last_val_secon
d_last_row)
as follows:
Figure 1.35: Use of comma-separated

notation
Slicing
7. To create a 2x2 matrix that starts at

the second row and second column,
we use [1:3, 1:3]:
# slicing an
intersection of 4
elements (2x2) of the
first two rows and
first two columns
subsection_2x2 =
dataset[1:3, 1:3]
np.mean(subsection_2x2
)

as follows:
Figure 1.36: A screenshot of the mean of

the 2x2 subset
8. Introducing the second column into
the indexing allows us to add
another layer of complexity. The
third value allows us to only select
certain values like every other
element by providing the value of
2. This means it skips the values
between and only takes each second
element from the used list. In this
task, we want to have every other
element, so we provide an indexing
of ::2, which, as we already
discussed, will take every other
element of the whole list:
# selecting every
second element of the
fifth row
every_other_elem =
dataset[6, ::2]
np.mean(every_other_el
em)

as follows:
Figure 1.37: Selecting every other

element of the seventh row
9. Negative numbers can also be used

to reverse the elements in a slice:
# reversing the entry

order, selecting the
first two rows in
reversed order
reversed_last_row =
dataset[-1, ::-1]
np.mean(reversed_last_
row)

as follows:
Figure 1.38: Slice of the last row with the elements in

reversed order
Splitting
1. Horizontally splitting our data can

be done with the hsplit method.
Note that if the dataset can't be split
with the given number of slices, it
will throw an error:
# splitting up our
dataset horizontally
on indices one third
and two thirds
hor_splits =
np.hsplit(dataset,(3))
2. We now need to split the first third

into two equal parts vertically.
There is also a vsplit method
that does exactly this; it works the
same as with hsplit:
# splitting up our
dataset vertically on
index 2
ver_splits =
np.vsplit(hor_splits[0
],(2))
3. When comparing the shapes, we

can see that the subset has the
required half of rows and the third
half of columns:
# requested subsection
of our dataset which
has only half the
amount of rows and
only a third of the
columns
print("Dataset",
dataset.shape)
print("Subset",
ver_splits[0].shape)

as follows:
Figure 1.39: Comparing the shapes of

the original dataset and subset
Iterating
4. Looking at the given piece of code,

we can see that the index is simply
incremented with each element.
This only works with one-

dimensional data. If we want to
index multi-dimensional data, this
won't work:
# iterating over whole

dataset (each value in
each row)
curr_index = 0
for x in
np.nditer(dataset):
print(x, curr_index)
curr_index += 1

as follows:
Figure 1.40: Iterating the entire dataset
5. ndenumerate is the right method

to use for this task. In addition to
the value, it returns the index. This
works with multi-dimensional data,
too:
# iterating over whole

dataset with indices
matching the position
in the dataset
for index, value in

np.ndenumerate(dataset
):
print(index, value)

as follows:
Figure 1.41: Enumerating the dataset with multi-
dimensional data

wrangling methods for NumPy. In the next activity, we'll take a
look at more advanced features that will give you the tools to
get better insights into your data.
SORTING, COMBINING,
AND RESHAPING
Solution:
Let's use the filtering features of NumPy for sorting, stacking,

combining, and reshaping our data:
# importing the
import numpy as np
2. Load the
look at the ndarray, like in the
previous activity:
dataset =
v', delimiter=',')
Filtering
3. Getting values greater than 105 can

be done by supplying the condition
in the brackets:
# values that are

greater than 105
vals_greater_five =
dataset[dataset > 105]
4. To use more complex conditions,

we might want to use the extract
method of NumPy. However, we
can also have the same checks with
the bracket notation:
# values that are
between 90 and 95
vals_between_90_95 =
np.extract((dataset >
90) & (dataset < 95),
dataset)
5. The where method of numpy

allows us to only get indices
(rows, cols) for each of the
matching values. In this task, we
want to print them out nicely. We
can combine rows with the
respective cols using List
comprehension. In this
example, we have simply added the
column to the respective row:
# indices of values
that have a delta of
less than 1 to 100
rows, cols =
np.where(abs(dataset -
100) < 1)
one_away_indices =
[[rows[index],
cols[index]] for
(index, _) in
np.ndenumerate(rows)]
Note
List comprehensions are Python's

way of mapping over data. They're
a handy notation for creating a new
list with some operation applied to
every element of the old list.
For example, if we want to double

the value of every element in this
list, list = [1, 2, 3, 4,
5], we would use list
comprehensions like
this:doubled_list=[x*x for
x in list]. This would give us
the following list: [1, 4, 9,
16, 25]. To get a better
understanding of list
comprehensions, please visit
https://docs.python.org/3/tutorial/d
atastructures.html#list-
comprehensions.
Sorting
6. Sorting each row in our dataset

works by using the sort method.
As described in the notebook, this
will always take the last axis, which
in this case is the right one to sort
per row:
# values sorted for
each row
row_sorted =
np.sort(dataset)
7. With multi-dimensional data, we

can use the axis parameter to
define which dataset should be
sorted. 0, in this case, means
column-based sorting:
# values sorted for

each column
col_sorted =
np.sort(dataset,
axis=0)
8. If we want to keep the order of our

dataset and only want to know
which indices the values in a sorted
dataset would have, we can use
argsort. In combination with
fancy indexing, we can get access
to sorted elements easily:
# indices of positions
for each row
index_sorted =
np.argsort(dataset)
Combining
9. Use combining features to add the
second half of the first column back
together, add the second column to
our combined dataset, and add the
third column to our combined
dataset.
# split up dataset
from activity03
thirds =
np.hsplit(dataset,
(3))
halfed_first =
np.vsplit(thirds[0],
(2))
# this is the part

we've sent the client
in activity03
halfed_first[0]

as follows:
Figure 1.42: Splitting the dataset
10. Depending on which way we want

to combine our data, we have to
either use vstack or hstack.
Both take a list of datasets to stack
together:
# adding the second

half of the first
column to the data
first_col =
np.vstack([halfed_firs
t[0],
halfed_first[1]])
11. After vstacking the second half of

our split dataset, we have one third
of our initial dataset stacked
together again. We now want to add
the other two remaining datasets to
our first_col dataset. We can
do this by using the hstack
method, which combines our
already combined first_col
with the second of the three split
datasets:
# adding the second

column to our combined
dataset
first_second_col =
np.hstack([first_col,
thirds[1]])
12. To reassemble our initial dataset,

one third is still missing. We can
hstack the last one-third column
onto our dataset, which is the same
thing we did with our second-third
column in the previous step:
# adding the third

column to our combined
dataset
full_data =
np.hstack([first_secon
d_col, thirds[2]])
Reshaping
13. The first subtask is to reshape our

dataset into a single list. This is
done using the reshape method:
# reshaping to a list
of values
single_list =
np.reshape(dataset,
(1, -1))
14. If we supply a -1 for the dimension

we don't know about, NumPy
figures this dimension out itself:
# reshaping to a
matrix with two
columns
two_col_dataset =
dataset.reshape(-1, 2)
ACTIVITY 4: USING
PANDAS TO COMPUTE
VARIANCE FOR THE
GIVEN NUMBERS
Solution:
Let's use pandas' features such as mean, median, and variance

to make some calculations on our data:

# importing the
import pandas as pd

aforementioned dataset. We want to
use the first column, containing the
parameter for that. The full line
should look like this:
dataset =
index_col=0)
3. First, we want to print a subset of

the dataset with the first two rows.
We can, again, use the Python List
syntax to create a subset of the
DataFrame of the first two rows:
# looking at the first

two rows of the
dataset
dataset[0:2]

as follows:
Figure 1.43: The first two rows, printed
4. Once we know that our dataset has

been successfully loaded, we can
start solving the given tasks.
The third row can be accessed by

indexing dataset.iloc[[2]].
To get the mean of the country
rather than the yearly column, we
need to pass the axis parameter:

of the third row
dataset.iloc[[2]].mean
(axis=1)

as follows:

third row
5. The last element of a DataFrame

can, just like with NumPy ndarrays
and Python Lists, be indexed using
-1 as the index. So,
dataset.iloc[[-1]] will
give us the last row:

of the last row
dataset.iloc[[-1]].mea
n(axis=1)

as follows:
last row
6. Besides using iloc to access rows

based on their index, we can also
use loc, which works based on the
index column. This can be defined
by using index_col=0 in the
read_csv call:

of the country Germany
dataset.loc[["Germany"
]].mean(axis=1)

as follows:
Figure 1.46: Indexing a country and

calculating the mean of Germany
7. Moving to the next set of tasks,

which cover the usage of the
median, we will see that the API is
consistent between the different
methods. This means that we have
to keep providing axis=1 to our
method calls to make sure that we
are aggregating for each country:

of the last row
dataset.iloc[[-1]].med
ian(axis=1)

as follows:
Figure 1.47: Usage of the median

method on the last row
8. Slicing rows in pandas is similar to

doing so in NumPy. We can use
reverse indexing to get the last
three columns with
dataset[-3:]:

of the last 3 rows
dataset[-3:].median(ax
is=1)

as follows:
Figure 1.48: Median of the last three
columns
9. When handling larger datasets, the

order in which methods are
executed matters. Think about what
head(10) does for a moment. It
simply takes your dataset and
returns the first 10 rows in it,
cutting down your input to the
mean() method drastically.
This will definitely have an impact

when using more memory-intensive
calculations, so keep an eye on the
order:
of the first 10
countries
dataset.head(10).media
n(axis=1)

as follows:
Figure 1.49: Usage of axis to calculate
the median of the first 10 rows
10. The last method we'll cover here is

the variance. Again, pandas
provides us with a consistent API,
which makes its usage easy. Since
we only want to display the last five
columns, we will make use of the
tail method:
# calculate the
variance of the last 5
columns
dataset.var().tail()

as follows:
Figure 1.50: Variance among the last five
columns
11. As mentioned in the introduction to

pandas, it's interoperable with
several features of NumPy.
Here's an example of how to use

NumPy's mean method with a
pandas DataFrame. In some cases,
NumPy has better functionality but
pandas, with its DataFrames, has a
better format:
# NumPy pandas
interoperability
import numpy as np
print("pandas",
)
print("numpy",
np.mean(dataset["2015"
]))

as follows:
Figure 1.51: Using NumPy's mean method with a

pandas DataFrame
Congratulations! You've completed your first activity with

pandas, which showed you some of the similarities but also
differences when working with NumPy and pandas. In the
following activities, this knowledge will be consolidated.
You'll also be introduced to more complex features and
methods of pandas.
USING PANDAS
Solution:
Let's use the indexing, slicing, and iterating operations to

display the population density of Germany, Singapore, United
States, and India for years 1970, 1990, and 2010.
Indexing
# importing the
import pandas as pd

mentioned dataset. We want to use
the first column, containing the
dataset =
index_col=0)
3. To index the row with the

index_col "United States", we
need to use the loc method:
# indexing the USA row
dataset.loc[["United
States"]].head()

as follows:
Figure 1.52: Indexing United States with
the loc method
4. Reverse indexing can also be used

with pandas. To index the second to
last row, we need to use the iloc
method, which accepts int-type
data, meaning an index:
# indexing the last

second to last row by
index
dataset.iloc[[-2]]

as follows:
Figure 1.53: Indexing the second to last
row
5. Columns are indexed using their

header. This is the first line of the
CSV file. To get the column with
the header 2000, we can use
normal indexing. Remember, the
head() method simply returns the
first five rows:
# indexing the column

of 2000 as a Series
dataset["2000"].head()

as follows:
Figure 1.54: Indexing all 2,000 columns
6. Since the double brackets notation

again returns a DataFrame, we can
chain method calls to get distinct
elements. To get the population
density of India in 2000, we first
want to get the data for the year
2000 and only then select India
using the loc() method:
# indexing the
population density of
India in 2000
(Dataframe)
dataset[["2000"]].loc[
["India"]]

as follows:
Figure 1.55: Getting the population

density of India in 2000
7. If we want to only retrieve a Series

object, we have to replace the
double brackets with single ones.
This will give us the distinct value
instead of the new DataFrame:
# indexing the
population density of
India in 2000 (Series)
dataset["2000"].loc["I
ndia"]

as follows:
Figure 1.56: India's population density

in the year 2000
Slicing
8. To create a slice with the rows 2 to

5, we have to use the iloc()
method again. We can simply
provide the same slicing syntax as
NumPy:
# slicing countries of
rows 2 to 5
dataset.iloc[1:5]
as follows:
Figure 1.57: The countries in rows 2 to 5
9. Using the loc() method, we can

also access dedicated rows by their
index_col (which was defined
in the read_csv call). To get
several rows in a new DataFrame,
we can use the nested brackets to
provide a list of elements:
# slicing rows
Germany, Singapore,
United States, and
India
dataset.loc[["Germany"
, "Singapore", "United
States", "India"]]

as follows:
Figure 1.58: Slicing countries Germany,
Singapore, United States, and India
10. Since the double bracket queries

return new DataFrames, we can
chain methods and therefore access
distinct subframes of our data:
# slicing a subset of
Germany, Singapore,
United States, and
India
# for years 1970,

1990, 2010 <
country_list =
["Germany",
"Singapore", "United
States", "India"]
dataset.loc[country_li
st][["1970", "1990",
"2010"]]

as follows:
Figure 1.59: Slices of Germany,
Singapore, United States, and India with
their population density of years 1970,
1990, and 2010
Iterating
11. To iterate our dataset and print out

the countries up until Angola, we
can use the iterrows() method.
The index will be the name of our
row, and the row will hold all
columns:
# iterating over the

first three countries
(row by row)
for index, row in

dataset.iterrows():
# only printing the

rows until Angola
if index == 'Angola':
break
print(index, '\n',
row[["Country Code",
"1970", "1990",
"2010"]], '\n')

as follows:
Figure 1.60: Iterating all countries until Angola

wrangling methods using pandas. In the next activity, we'll take
a look at more advanced features such as filtering, sorting, and
reshaping to prepare you for the next chapter.
SORTING, AND
RESHAPING
Solution:
Let's use pandas to filter, sort, and reshape our data.
Filtering
# importing the
import pandas as pd

aforementioned dataset. We want to
use the first column, containing the
dataset =
index_col=0)
3. Instead of using the bracket syntax,

we can also use the filter
method to filter for specific items.
Here, we also provide a list of
elements that should be kept:
# filtering columns
1961, 2000, and 2015
dataset.filter(items=
["1961", "2000",
"2015"]).head()

as follows:
Figure 1.61: Filtering data for the years

1961, 2000, and 2015
4. If we want to filter for specific

values in a specific column, we can
use conditions. To get all countries
that had a higher population density
than 500 in 2000, we simply pass
this condition in brackets:
# filtering countries
that had a greater
population density
than 500 in 2000
dataset[(dataset["2000
"] > 500)][["2000"]]

as follows:
Figure 1.62: Filtering out values that are
greater than 500 in the 2000 column
5. One powerful parameter of the

filter method is regex. This
allows us to search for arbitrary
columns or rows (depending on the
index given) that match a certain
regex. To get all the columns that
start with 2, we can simply pass ^2,
which means it starts with 2:
# filtering for years

2000 and later
dataset.filter(regex="
^2", axis=1).head()

as follows:
Figure 1.63: Retrieving all columns

starting with 2
6. Using the axis parameter, we can

decide on which dimension the
filtering should happen. To filter the
rows instead of the columns, we
can pass axis=0. This will be
helpful for situations like when we
want to filter all the rows that start
with A:
that start with A
dataset.filter(regex="
^A", axis=0).head()

as follows:
Figure 1.64: Retrieving the rows that
start with A
7. If we want all the rows or columns

that contain some specific value or
character, we can use the like
query. For example, if we want to
have only countries that contain the
word "land", such as Switzerland:
that contain the word
land
dataset.filter(like="l
and", axis=0).head()

as follows:
Figure 1.65: Retrieving all countries
containing the word "land"
Sorting
8. Sorting in pandas can be done by

using the sort_values or
sort_index methods. If we want
to get the countries with the lowest
population density for a specific
year, we can sort by this specific
column:
# values sorted by
column 1961
dataset.sort_values(by
=["1961"])
[["1961"]].head(10)

as follows:
Figure 1.66: Sorting by the values for
the year 1961
9. Just for comparison, we want to do

the same sorting for 2015. This
will give us some nice insights into
our data. We can see that the order
of the countries with the lowest
population density changed a bit,
but that the first three entries are
unchanged:
# values sorted by
column 2015
=["2015"])
[["2015"]].head(10)

as follows:
Figure 1.67: Sorting based on the values
of 2015
10. The default sorting order is

ascending. This means that we
have to provide a separate
parameter if we want to sort in
descending order, showing us
the biggest values first:
# values sorted by
column 2015 in
descending order
=["2015"],
ascending=False)
[["2015"]].head(10)

as follows:
Figure 1.68: Sorting in descending order
Reshaping
11. As we mentioned before, reshaping

data with pandas can be a very
complex task, so we'll only do an
easier reshape. We want to get
DataFrames where the columns are
country codes and the only
row is the year 2015. Since we
only have one 2015 label, we need
to duplicate it as many times as our
dataset length. This will lead to
every value receiving the 2015 row
index:
# reshaping to 2015 as
row and country codes
as columns
dataset_2015 =
dataset[["Country
Code", "2015"]]
dataset_2015.pivot(ind
ex=["2015"] *
len(dataset_2015),
columns="Country
Code", values="2015")

as follows:
Figure 1.69: Reshaping the dataset into a single row
for the values of 2015
You've now completed the topic about pandas, which

concludes this chapter. We've learned about the basic tools that
help you wrangle and work with data. pandas itself is an
incredibly powerful and heavily used tool for data wrangling
and understanding.
Chapter 2: All You
Need to Know about
Plots
SKILL COMPARISON
Solution:
1. Bar charts and radar charts are great

for comparing multiple variables
for multiple groups.
2. Suggested response: The bar chart

is great for comparing the skill
attributes of the different
employees, but it is
disadvantageous to get an overall
impression for an employee. The
radar chart is great for comparing
values across both employees and
skill attributes.
3. Suggested response:
Bar chart: Addition of a title and

labels, use of different colors
Radar chart: Addition of a title and
different colors for different
employees
ACTIVITY 8: ROAD
ACCIDENTS OCCURRING
OVER TWO DECADES
Solution:
1. Suggested response: If we look at

the Figure 2.20, we can see that for
the year 2015, the number of
accidents have reduced to 100 in
the months of January and July.
Around 300 accidents have
occurred in the months of April and
October.

the trend for each month, that is,
January, April, July, and October
for the past two decades, we can
see a decreasing trend in the
number of accidents taking place in
January.
Design practices:
Select colors and contrasts that will

be easily visible to individuals with
vision problems so that your plots
are more inclusive.
ACTIVITY 9:
SMARTPHONE SALES
UNITS
Solution:
1. Suggested response: If we
compare the performance of each
manufacturer in the third and fourth
quarters, we come to the conclusion
that Apple has performed
exceptionally well. Their sales units
have risen at a higher rate from the
third quarter to the fourth quarter
for both 2016 and 2017, when
compared with that of other
manufacturers.

the trends in the sales units of each
manufacturer, we can see that after
the third quarter of 2017, the sales
units of all the companies except
Xiaomi have shown an
inconsistency. If we look at the
performance of Xiaomi, there has
been an upward trend after the first
quarter of the year 2017.

OF TRAINS DURING
DIFFERENT TIME
INTERVALS
Solution:
1. Suggested response: If we focus

on the preceding histogram, we can
clearly identify that most trains
arrive during 6 pm and 8 pm.
2. Suggested response: The

histogram will look as follows if
the number of trains between 4-6
pm are increased by 50:
Figure 2.45: Frequency of trains increased by 50
between 4-6 pm

THE IDEAL
VISUALIZATION
Solution:
First visualization: suggested response:

1. The proposed visualization has
multiple faults: first, a pie chart is
supposed to show part-of-a-whole
relations, which is not the case for
this task since we only consider the
top 30 YouTubers. Second, 30
values are too many to visualize
within a pie chart. Third, the slices
are not ordered according to their
size. Also, it is difficult to quantify
the slices as there is no unit of
measurement specified. On the
other hand, in the following
horizontal bar chart, it is easier to
tell the number of subscribers in
millions for each YouTube channel:
Figure 2.46: Horizontal bar chart showing Top 30
YouTubers
Second visualization: suggested response:
1. This is also an example of using the

wrong chart type. A line chart was
used to compare different
categories that do not have any
temporal relation. Furthermore,
information such as legends and
labels are missing. The following
figure shows how the data should
had been represented using
comparative bar chart:
Figure 2.47: Comparative bar chart displaying casino
data for two days
Chapter 3: A Deep
Dive into Matplotlib
STOCK TRENDS BY
USING A LINE PLOT
Solution:
Let's visualize a stock trend by using a line plot:
1. Open the
activity12_solution.ipyn
this activity.

command-line terminal:
jupyter-lab

Notebook:
# Import statements
import
plt
import numpy as np
import pandas as pd
%matplotlib inline

in the data folder. The
read_csv() function reads a .csv
file into DataFrame:
# load datasets
google =
pd.read_csv('./data/GO
OGL_data.csv')
facebook =
pd.read_csv('./data/FB
_data.csv')
apple =
pd.read_csv('./data/AA
PL_data.csv')
amazon =
pd.read_csv('./data/AM
ZN_data.csv')
microsoft =
pd.read_csv('./data/MS
FT_data.csv')
4. Use Matplotlib to create a line
chart that visualizes the closing
prices for the past five years (whole
data sequence) for all five
companies. Add labels, titles, and a
legend to make the visualization
self-explanatory. Use the
plt.grid() function to add a
grid to your plot:
# Create figure
plt.figure(figsize=
(16, 8), dpi=300)
# Plot data
plt.plot('date',
'close', data=google,
label='Google')
plt.plot('date',
'close',
data=facebook,
label='Facebook')
plt.plot('date',
'close', data=apple,
label='Apple')
plt.plot('date',
'close', data=amazon,
label='Amazon')
plt.plot('date',
'close',
data=microsoft,
label='Microsoft')
# Specify ticks for x

and y axis
plt.xticks(np.arange(0
, 1260, 40),
rotation=70)
plt.yticks(np.arange(0
, 1450, 100))
# Add title and label

for y-axis
plt.title('Stock
trend', fontsize=16)
plt.ylabel('Closing
price in $',
fontsize=14)
# Add grid
plt.grid()
# Add legend
plt.legend()
# Show plot
plt.show()

BAR PLOT FOR MOVIE
COMPARISON
Solution:
Let's create a bar plot for comparing the ratings of different

movies:
1. Open the
this activity.

jupyter-lab.

Notebook.;
# Import statements
import numpy as np
import pandas as pd
import
plt
%matplotlib inline
in the data folder:
# Load dataset
movie_scores =
pd.read_csv('./data/mo
vie_scores.csv')
4. Use Matplotlib to create a visually

appealing bar plot comparing the
two scores for all five movies. Use
the movie titles as labels for the x-
axis. Use percentages in an interval
of 20 for the y-axis and minor ticks
in an interval of 5. Add a legend
and a suitable title to the plot:
# Create figure
plt.figure(figsize=
(10, 5), dpi=300)
# Create bar plot
pos =
np.arange(len(movie_sc
ores['MovieTitle']))
width = 0.3
plt.bar(pos - width /
2,
movie_scores['Tomatome
ter'], width,
label='Tomatometer')
plt.bar(pos + width /
2,
movie_scores['Audience
Score'], width,
label='Audience
Score')
# Specify ticks
plt.xticks(pos,
rotation=10)
plt.yticks(np.arange(0
, 101, 20))
# Get current Axes for

setting tick labels
and horizontal grid
ax = plt.gca()
# Set tick labels
ax.set_xticklabels(mov
ie_scores['MovieTitle'
])
ax.set_yticklabels(['0
%', '20%', '40%',
'60%', '80%', '100%'])
# Add minor ticks for

y-axis in the interval
of 5
ax.set_yticks(np.arang
e(0, 100, 5),
minor=True)
# Add major horizontal

grid with solid lines
ax.yaxis.grid(which='m
ajor')
# Add minor horizontal

grid with dashed lines
ax.yaxis.grid(which='m
inor', linestyle='--')
# Add title
plt.title('Movie
comparison')
# Add legend
plt.legend()
# Show plot
plt.show()
5. Some functions require you to

explicitly specify the axes. To get
the reference to the current axes,
use ax = plt.gca(). Ticks can
be specified using
plt.xticks([xticks]) and
plt.yticks([yticks]) for
the x-axis and y-axis, respectively.
Axes.set_xticklabels([la
bels]) and
Axes.set_yticklabels([la
bels]) can be used to set tick
labels. To add minor y-ticks, use
Axes.set_yticks([ticks],
minor=True). To add a
horizontal grid for major ticks, use
Axes.yaxis.grid(which='m
ajor'), and to add a dashed
horizontal grid for minor ticks, use
Axes.yaxis.grid(which='m
inor', linestyle='--').

STACKED BAR PLOT TO
PERFORMANCE
Solution:
Let's create a stacked bar chart to visualize the performance of

a restaurant:
1. Open the
this activity.

jupyter-lab.

Notebook:
# Import statements
import pandas as sb
import numpy as np
import
plt
%matplotlib inline
3. Load the dataset:
# Load dataset
bills =
sns.load_dataset('tips
')
4. Use the given dataset and create a

matrix where the elements contain
the sum of the total bills for each
day and smokers/non-smokers:
days = ['Thur', 'Fri',

'Sat', 'Sun']
days_range =
np.arange(len(days))
smoker = ['Yes', 'No']
bills_by_days =
[bills[bills['day'] ==
day] for day in days]
bills_by_days_smoker =
[[bills_by_days[day]
[bills_by_days[day]
['smoker'] == s] for s
in smoker] for day in
days_range]
total_by_days_smoker =
[[bills_by_days_smoker
[day][s]
['total_bill'].sum()
for s in
range(len(smoker))]
for day in days_range]
totals =
np.asarray(total_by_da
ys_smoker)
Here, the asarray() function is

used to convert any given input into
an array.
5. Create a stacked bar plot, stacking

the summed total bills separated by
smoker and non-smoker for each
day. Add a legend, labels, and a
title:
# Create figure
plt.figure(figsize=
(10, 5), dpi=300)
# Create stacked bar

plot
plt.bar(days_range,
totals[:, 0],
label='Smoker')
plt.bar(days_range,
totals[:, 1],
bottom=totals[:, 0],
label='Non-smoker')
# Add legend
plt.legend()
# Add labels and title
plt.xticks(days_range)
ax = plt.gca()
ax.set_xticklabels(day
s)
ax.yaxis.grid()
plt.ylabel('Daily
total sales in $')
plt.title('Restaurant
performance')
# Show plot
plt.show()

SMARTPHONE SALES
AREA CHART
Solution:
Let's compare the sales units of smartphone manufacturers

using a stacked area chart:
1. Open the
this activity.

jupyter-lab.
Notebook:
# Import statements
import pandas as pd
import numpy as np
import
plt
%matplotlib inline

in the data folder:
# Load dataset
sales =
pd.read_csv('./data/sm
artphone_sales.csv')
4. Create a visually appealing stacked

area chart. Add a legend, labels,
and a title:
# Create figure
plt.figure(figsize=
(10, 6), dpi=300)
# Create stacked area

chart
labels =
sales.columns[1:]
plt.stackplot('Quarter
', 'Apple', 'Samsung',
'Huawei', 'Xiaomi',
'OPPO', data=sales,
labels=labels)
# Add legend
plt.legend()
plt.xlabel('Quarters')
plt.ylabel('Sales
units in thousands')
plt.title('Smartphone
sales units')
# Show plot
plt.show()

HISTOGRAM AND A BOX
Solution:
Let's visualize the intelligence quotient of different groups

using a histogram and a box plot:
1. Open the
this activity.

jupyter-lab.

Notebook:
# Import statements
import numpy as np
import
plt
%matplotlib inline
3. Use the following code to generate

a sample of IQ scores:
# IQ samples
iq_scores = [126, 89,

90, 101, 102, 74, 93,
101, 66, 120, 108, 97,
98, 105, 119, 92, 113,
81, 104, 108, 83, 102,
105, 111, 102, 107,
103, 89, 89, 110, 71,
110, 120, 85, 111, 83,
122, 120, 102, 84,
118, 100, 100, 114,
81, 109, 69, 97, 95,
106, 116, 109, 114,
98, 90, 92, 98, 91,
81, 85, 86, 102, 93,
112, 76, 89, 110, 75,
100, 90, 96, 94, 107,
108, 95, 96, 96, 114,
93, 95, 117, 141, 115,
95, 86, 100, 121, 103,
66, 99, 96, 111, 110,
105, 110, 91, 112,
102, 112, 75]
4. Plot a histogram with 10 bins for

the given IQ scores. IQ scores are
normally distributed with a mean of
100 and a standard deviation of 15.
Visualize the mean as a vertical
solid red line, and the standard
deviation using dashed vertical
lines. Add labels and a title:
# Create figure
4), dpi=150)
# Create histogram
plt.hist(iq_scores,
bins=10)
plt.axvline(x=100,
color='r')
plt.axvline(x=115,
color='r', linestyle=
'--')
plt.axvline(x=85,
color='r', linestyle=
'--')
plt.xlabel('IQ score')
plt.ylabel('Frequency'
)
plt.title('IQ scores
for a test group of a
hundred adults')
# Show plot
plt.show()
5. Create a box plot to visualize the

same IQ scores. Add labels and a
title:
# Create figure
4), dpi=150)
# Create histogram
plt.boxplot(iq_scores)
ax = plt.gca()
ax.set_xticklabels(['T
est group'])
plt.ylabel('IQ score')
for a test group of a
hundred adults')
# Show plot
plt.show()
6. The following are IQ scores for

different test groups:
group_a = [118, 103,

125, 107, 111, 96,
104, 97, 96, 114, 96,
75, 114,
107, 87, 117, 117,

114, 117, 112, 107,
133, 94, 91, 118, 110,
117, 86, 143, 83, 106,

86, 98, 126, 109, 91,
112, 120, 108,
111, 107, 98, 89, 113,

117, 81, 113, 112, 84,
115, 96, 93,
128, 115, 138, 121,

87, 112, 110, 79, 100,
84, 115, 93, 108,
130, 107, 106, 106,

101, 117, 93, 94, 103,
112, 98, 103, 70,
139, 94, 110, 105,

122, 94, 94, 105, 129,
110, 112, 97, 109,
121, 106, 118, 131,

88, 122, 125, 93, 78]
group_b = [126, 89,

90, 101, 102, 74, 93,
101, 66, 120, 108, 97,
98,
105, 119, 92, 113, 81,

104, 108, 83, 102,
105, 111, 102, 107,
103, 89, 89, 110, 71,

110, 120, 85, 111, 83,
122, 120, 102,
84, 118, 100, 100,

114, 81, 109, 69, 97,
95, 106, 116, 109,
114, 98, 90, 92, 98,

91, 81, 85, 86, 102,
93, 112, 76,
89, 110, 75, 100, 90,

96, 94, 107, 108, 95,
96, 96, 114,
93, 95, 117, 141, 115,

95, 86, 100, 121, 103,
66, 99, 96,
111, 110, 105, 110,

91, 112, 102, 112, 75]
group_c = [108, 89,

114, 116, 126, 104,
113, 96, 69, 121, 109,
102, 107,
122, 104, 107, 108,

137, 107, 116, 98,
132, 108, 114, 82, 93,
89, 90, 86, 91, 99,

98, 83, 93, 114, 96,
95, 113, 103,
81, 107, 85, 116, 85,

107, 125, 126, 123,
122, 124, 115, 114,
93, 93, 114, 107, 107,

84, 131, 91, 108, 127,
112, 106, 115,
82, 90, 117, 108, 115,

113, 108, 104, 103,
90, 110, 114, 92,
101, 72, 109, 94, 122,

90, 102, 86, 119, 103,
110, 96, 90,
110, 96, 69, 85, 102,

69, 96, 101, 90]
group_d = [ 93, 99,

91, 110, 80, 113, 111,
115, 98, 74, 96, 80,
83,
102, 60, 91, 82, 90,

97, 101, 89, 89, 117,
91, 104, 104,
102, 128, 106, 111,

79, 92, 97, 101, 106,
110, 93, 93, 106,
108, 85, 83, 108, 94,

79, 87, 113, 112, 111,
111, 79, 116,
104, 84, 116, 111,

103, 103, 112, 68, 54,
80, 86, 119, 81,
84, 91, 96, 116, 125,

99, 58, 102, 77, 98,
100, 90, 106,
109, 114, 102, 102,

112, 103, 98, 96, 85,
97, 110, 131, 92,
79, 115, 122, 95, 105,

74, 85, 85, 95]

scores of different test groups. Add
labels and a title:
# Create figure
4), dpi=150)
# Create histogram
plt.boxplot([group_a,
group_b, group_c,
group_d])
ax = plt.gca()
ax.set_xticklabels(['G
roup A', 'Group B',
'Group C', 'Group D'])
plt.ylabel('IQ score')
for different test
groups')
# Show plot
plt.show()

SCATTER PLOT TO
BETWEEN VARIOUS
ANIMALS
Solution:
Let's visualize correlation between various animals with the

help of a scatter plot:
1. Open the
this activity.

jupyter-lab.

Notebook:
# Import statements
import pandas as pd
import numpy as np
import
plt
%matplotlib inline

in the data folder:
# Load dataset
data =
pd.read_csv('./data/an
age_data.csv')

Sort the data according to the
animal class. Here, the
isfinite() function checks for
the finiteness of the given element:
# Preprocessing
longevity = 'Maximum
longevity (yrs)'
mass = 'Body mass (g)'
data =
data[np.isfinite(data[
longevity]) &
np.isfinite(data[mass]
)]
# Sort according to
class
amphibia =
data[data['Class'] ==
'Amphibia']
aves =
'Aves']
mammalia =
'Mammalia']
reptilia =
'Reptilia']
5. Create a scatter plot visualizing the

correlation between the body mass
and the maximum longevity. Use
different colors for grouping data
samples according to their class.
Add a legend, labels, and a title.
Use a log scale for both the x-axis
and y-axis:
# Create figure
plt.figure(figsize=
(10, 6), dpi=300)
# Create scatter plot
plt.scatter(amphibia[m
ass],
amphibia[longevity],
label='Amphibia')
plt.scatter(aves[mass]
, aves[longevity],
label='Aves')
plt.scatter(mammalia[m
ass],
mammalia[longevity],
label='Mammalia')
plt.scatter(reptilia[m
ass],
reptilia[longevity],
label='Reptilia')
# Add legend
plt.legend()
# Log scale
ax = plt.gca()
ax.set_xscale('log')
ax.set_yscale('log')
# Add labels
plt.xlabel('Body mass
in grams')
plt.ylabel('Maximum
longevity in years')
# Show plot
plt.show()

SCATTER PLOT WITH
MARGINAL HISTOGRAMS
Solution:
1. Open the
this activity.

jupyter-lab.

Notebook:
# Import statements
import pandas as pd
import numpy as np
import
plt
%matplotlib inline

in the data folder:
# Load dataset
data =
pd.read_csv('./data/an
age_data.csv')

Select all the samples of the aves
class and with a body mass smaller
than 20,000:
# Preprocessing
longevity (yrs)'
data =
longevity]) &
)]
# Sort according to
class
aves =
'Aves']
aves = data[data[mass]
< 20000]
5. Create a figure with a constrained

layout. Create a gridspec of size
4x4. Create a scatter plot of size
3x3 and marginal histograms of
size 1x3 and 3x1. Add labels and a
figure title:
# Create figure
fig =
8), dpi=150,
constrained_layout=Tru
e)
# Create gridspec
gs =
fig.add_gridspec(4, 4)
# Specify subplots
histx_ax =
fig.add_subplot(gs[0,
:-1])
histy_ax =
fig.add_subplot(gs[1:,
-1])
scatter_ax =
fig.add_subplot(gs[1:,
:-1])
# Create plots
scatter_ax.scatter(ave
s[mass],
aves[longevity])
histx_ax.hist(aves[mas
s], bins=20,
density=True)
histx_ax.set_xticks([]
)
histy_ax.hist(aves[lon
gevity], bins=20,
density=True,
orientation='horizonta
l')
histy_ax.set_yticks([]
)

plt.xlabel('Body mass
in grams')
plt.ylabel('Maximum
longevity in years')
fig.suptitle('Scatter
plot with marginal
histograms')
# Show plot
plt.show()

GRID
Solution:
1. Open the
this activity.

jupyter-lab.

Notebook:
# Import statements
import os
import numpy as np
import
plt
import
matplotlib.image as
mpimg
%matplotlib inline
3. Load all four images from the data

folder:
# Load images
img_filenames =
os.listdir('data')
imgs =
[mpimg.imread(os.path.
join('data',
img_filename)) for
img_filename in
img_filenames]
4. Visualize the images in a 2x2 grid.

Remove the axes and give each
image a label:
# Create subplot
fig, axes =
plt.subplots(2, 2)
fig.figsize = (6, 6)
fig.dpi = 150
axes = axes.ravel()
# Specify labels
labels = ['coast',
'beach', 'building',
'city at night']
# Plot images
for i in
range(len(imgs)):
axes[i].imshow(imgs[i]
)
axes[i].set_xticks([])
axes[i].set_yticks([])
axes[i].set_xlabel(lab
els[i])
Chapter 4: Simplifying
Seaborn
IQ SCORES FOR
DIFFERENT TEST
PLOT
Solution:
Let's compare IQ scores among different test groups using the

Seaborn library:
1. Open the
this activity. Navigate to the path of
this file and type in the following at
the command-line terminal:
jupyter-lab.

Notebook.
%matplotlib inline
import numpy as np
import pandas as pd
import
plt
3. Use the pandas read_csv()

function to read the data located in
the data folder:
mydata =
pd.read_csv("./data/sc
ores.csv")
4. Access the data of each test group

in the column. Convert them into a
list using the tolist() method.
Once the data of each test group has
been converted into a list, assign
this list to variables of each
respective test groups:
group_a =
mydata[mydata.columns[
0]].tolist()
group_b =
1]].tolist()
group_c =
2]].tolist()
group_d =
3]].tolist()
5. Print the variables of each group to

check whether the data inside it is
converted into a list. This can be
done with the help of the print()
function:
print(group_a)
Data values of Group A are shown

in the following figure:
Figure 4.35: Values of Group A
print(group_b)
Data values of Group B are shown

Figure 4.36: Values of Group B
print(group_c)
Data values of Group C are shown
Figure 4.37: Values of Group C
print(group_d)
Data values of Group D are shown

Figure 4.38: Values of Group D
6. Once we have the data for each test

group, we need to construct a
DataFrame from this given data.
This can be done with the help of
the pd.DataFrame() function,
which is provided by pandas.
data =
pd.DataFrame({'Groups'
: ['Group A'] *
len(group_a) + ['Group
B'] * len(group_b) +
['Group C'] *
len(group_c) + ['Group
D'] * len(group_d),
'IQ score': group_a +

group_b + group_c +
group_d})
7. Now, since we have the DataFrame,
we need to create a box plot using
the boxplot() function that's
provided by Seaborn. Within this
function, we need to specify the
titles for both the axes along with
the DataFrame we are using. The
title for the x-axis would be
Groups and the title for the y-axis
would be IQ score. As far as the
DataFrame is concerned, we will
pass data as a parameter. Here,
data is the DataFrame that we
obtained from the previous step.
plt.figure(dpi=150)
# Set style
sns.set_style('whitegr
id')
# Create boxplot
sns.boxplot('Groups',
'IQ score', data=data)
# Despine
sns.despine(left=True,
right=True, top=True)
# Add title
for different test
groups')
# Show plot
plt.show()
The despine() function helps in

removing the top and right spines
from the plot. Here, we have also
removed the left spine. Using the
title() function, we have set the
title for our plot. The show()
function helps to visualize the plot.
From Figure 4.8, we can conclude that by using a box plot, the
IQ scores of Group A are better than the other groups.
ACTIVITY 21: USING

HEATMAPS TO FIND
PATTERNS IN FLIGHT
PASSENGERS' DATA
Solution:
Let's find the patterns in the flight passengers' data with the
help of a heatmap:
1. Open the
jupyter-lab.

Notebook:
%matplotlib inline
import numpy as np
import pandas as pd
import
plt
3. Use the read_csv() function of

pandas to read the data located in
the data folder:
mydata =
pd.read_csv("./data/fl
ight_details.csv")
4. Now, we can use the pivot()

function to provide meaningful row
and column labels to our
DataFrame:
data =
mydata.pivot("Months",
"Years", "Passengers")
5. Use the heatmap() function of

the Seaborn library to visualize this
data. Within this function, we pass
parameters such as DataFrame and
colormap. Since we got data from
the preceding code, we will pass it
as a DataFrame in the
heatmap() function. Also, we
will create our own colormap and
pass it as a second parameter to this
function.
sns.set()
plt.figure(dpi=150)
sns.heatmap(data,
cmap=sns.light_palette
("orange",
as_cmap=True,
reverse=True))
plt.title("Flight
Passengers from 2001
to 2012")
plt.show()
From Figure 4.23, we can conclude that the number of flight

passengers were highest in July 2012 and August 2012.
ACTIVITY 22: MOVIE

Solution:
Let's compare the movie scores for five different movies by

using a bar plot that's been provided by Seaborn library:
1. Open the
jupyter-lab.

Notebook:
%matplotlib inline
import numpy as np
import pandas as pd
import
plt

the data folder:
mydata =
pd.read_csv("./data/mo
vie_scores.csv")
4. Construct a DataFrame from this

given data. This can be done with
the help of the pd.DataFrame()
function, which is provided by
pandas. The following code gives
us a better idea of this:
movie_scores =
pd.DataFrame({"Movie
Title":
list(mydata["MovieTitl
e"]) * 2,
"Score":
list(mydata["AudienceS
core"]) +
list(mydata["Tomatomet
er"]),
"Type": ["Audience
Score"] *
len(mydata["AudienceSc
ore"]) +
["Tomatometer"] *
len(mydata["Tomatomete
r"])})
5. Make use of the barplot()

function provided by Seaborn.
Provide Movies and Scores as
parameters so that their data is
displayed on both axes. Provide
Type as hue, on the basis of
which the comparison needs to be
made. The last parameter needs a
DataFrame as input. Thus, we
provide the movie_scores
DataFrame, which we obtained
from the previous step.
The following code provides a

better understanding of this:
sns.set()
plt.figure(figsize=
(10, 5), dpi=300)
# Create bar plot
ax =
sns.barplot("Movie
Title", "Score",
hue="Type",
data=movie_scores)
plt.xticks(rotation=10
)
# Add title
plt.title("Movies
Scores comparison")
plt.xlabel("Movies")
plt.ylabel("Scores")
# Show plot
plt.show()
We compared the ratings of AudienceScore and Tomatometer

for 5 different movies and concluded that the ratings matched
for the movie The Martian.

IQ SCORES FOR
DIFFERENT TEST
GROUPS BY USING A
VIOLIN PLOT
Solution:
Let's compare IQ scores among different test groups using the

Seaborn library:
1. Open the
jupyter-lab.

Notebook:
%matplotlib inline
import numpy as np
import pandas as pd
import
plt

the data folder:
mydata =
pd.read_csv("./data/sc
ores.csv")

list using the tolist() method.
Once the data of each test group has
been converted into a list, assign
this list to the variables of each
respective test group.
group_a =
0]].tolist()
group_b =
1]].tolist()
group_c =
2]].tolist()
group_d =
3]].tolist()

check whether the data inside it has
been converted into a list. This can
be done with the help of the
print() function:
print(group_a)
Figure 4.39: Values of Group A
print(group_b)
Figure 4.40: Values of Group B
print(group_c)
Figure 4.41: Values of Group C
print(group_d)
Figure 4.42: Values of Group D
6. Once we get the data for each test

group, we need to construct a
DataFrame from this given data.
This can be done with the help of
the pd.DataFrame() function
that's provided by pandas.
data =
pd.DataFrame({'Groups'
: ['Group A'] *
len(group_a) + ['Group
B'] * len(group_b) +
['Group C'] *
len(group_c) + ['Group
D'] * len(group_d),
'IQ score': group_a +

group_b + group_c +
group_d})

we need to create a violin plot using
the violinplot() function
that's provided by Seaborn. Within
this function, we need to specify the
titles for both the axes along with
the DataFrame we are using. The
title for the x-axis will be Groups
and the title for the y-axis will be
IQ score. As far as the
DataFrame is concerned, we will
pass data as a parameter. Here,
data is the DataFrame that we
plt.figure(dpi=150)
# Set style
sns.set_style('whitegr
id')
# Create boxplot
sns.violinplot('Groups
', 'IQ score',
data=data)
# Despine
sns.despine(left=True,
right=True, top=True)
# Add title
for different test
groups')
# Show plot
plt.show()
The despine() function helps is

removing the top and right spines
from the plot. Here, we have also
removed the left spine. Using the
title() function, we have set the
title for our plot. The show()
function helps to visualize the plot.
We can conclude that the IQ scores of Group A are better than

those for other groups.
ACTIVITY 24: TOP 30

YOUTUBE CHANNELS
Solution:
Let's visualize the total number of subscribers and the total

number of views for the top 30 YouTube channels by using the
FacetGrid() function that's provided by the Seaborn
library:
1. Open the
jupyter-lab.

Notebook:
%matplotlib inline
import numpy as np
import pandas as pd
import
plt

the data folder:
mydata =
pd.read_csv("./data/yo
utube.csv")

list by using the tolist()
method. Once the data of each test
group has been converted into a list,
assign this list to variables of each
respective test group.
channels =
0]].tolist()
subs =
1]].tolist()
views =
2]].tolist()

check whether the data inside it has
been converted into a list. This can
be done with the help of the
print() function:
print(channels)
Figure 4.43: List of YouTube channels
print(subs)
Figure 4.44: List of Subscribers for each

YouTube channel
print(views)
Figure 4.45: List of Views for each

YouTube channel
6. Once we get the data for

channels, subs, and views,
we need to construct a DataFrame
from this given data. This can be
done with the help of the
pd.DataFrame() function that's
provided by pandas.
data =
pd.DataFrame({'YouTube
Channels': channels +
channels, 'Subscribers
in millions': subs +
views, 'Type':
['Subscribers'] *
len(subs) + ['Views']
* len(views)})

we need to create a FacetGrid using
the FacetGrid() function that's
provided by Seaborn. Here, data
is the DataFrame, which we
sns.set()
g =
sns.FacetGrid(data,
col='Type',
hue='Type',
sharex=False,
height=8)
g.map(sns.barplot,
'Subscribers in
millions', 'YouTube
Channels')
plt.show()
We can conclude that PewDiePie's YouTube channel has the

highest number of subscribers, whereas T-Series has the
highest number of views.
ACTIVITY 25: LINEAR

REGRESSION
Solution:
Let's visualize the linear relationship between maximum

longevity and body mass in the regression plot by using the
regplot() function that's provided by the Seaborn library:
1. Open the
jupyter-lab.

Notebook:
%matplotlib inline
import numpy as np
import pandas as pd
import
plt

the data folder:
mydata =
pd.read_csv("./data/an
age_data.csv")
4. Filter the data so that you end up

Only consider samples for the
Mammalia class and a body mass
below 200,000. This preprocessing
can be seen in the following code:
longevity (yrs)'
data =
mydata[mydata['Class']
== 'Mammalia']
data =
longevity]) &
) & (data[mass] <
200000)]
5. Once the preprocessing is done, we

need plot the data using the
regplot() function that's
provided by the Seaborn library. In
the following code, we have
provided three parameters inside
the regplot() function. The first
two parameters are mass and
longevity, wherein the body
mass data will be shown in the x-
axis and the maximum longevity
data will be shown in the y-axis. In
the third parameter, we need to
provide a DataFrame called data,
which we obtained from the
previous step:
# Create figure
sns.set()
plt.figure(figsize=
(10, 6), dpi=300)
# Create scatter plot

sns.regplot(mass,
longevity, data=data)
# Show plot
plt.show()
We can conclude that there is a linear relationship between

body mass and maximum longevity for the Mammalia class.
ACTIVITY 26: WATER

USAGE REVISITED
Solution:
Let's visualize the water usage by using a tree map, which can
be created with the help of the Squarify library:
1. Open the
Lesson 04 folder to implement
jupyter-lab.

Notebook:
%matplotlib inline
import numpy as np
import pandas as pd
import
plt
import squarify

the data folder:
mydata =
pd.read_csv("./data/wa
ter_usage.csv")
4. Create a list of labels by accessing

each column from the preceding
dataset. Here, the
astype('str') function is used
to cast the fetched data into a type
string.
# Create figure
plt.figure(dpi=200)
# Create tree map
labels =
mydata['Usage'] + ' ('
+
mydata['Percentage'].a
stype('str') + '%)'
5. For creating a tree map

visualization of given data, we
make use of the plot() function
of the squarify library. This
function takes three parameters.
The first parameter is the list of all
percentages and the second
parameter is the list of all labels,
which we got in the previous step.
The third parameter is the color that
can be created by using the
light_pallete() function of
the Seaborn library.
squarify.plot(sizes=my
data['Percentage'],
label=labels,
color=sns.light_palett
e('green',
mydata.shape[0]))
plt.axis('off')
# Add title
plt.title('Water
usage')
# Show plot
plt.show()
Chapter 5: Plotting
Geospatial Data
MAP
Solution:
Let's plot the geospatial data on a map and find the densely
populated areas for cities in Europe that have a population of
more than 100k:

this activity.
2. Before we can start working with

the data, we need to import the
dependencies:
# importing the
import numpy as np
import pandas as pd
import geoplotlib
3. Load the dataset using pandas:
#loading the Dataset

(make sure to have the
dataset downloaded)
Dataset =
rld_cities_pop.csv',
dtype = {'Region':
np.str})
Note
If we import our dataset without

defining the dtype of
the Region column as a String,
we will get a warning telling us that
it has a mixed datatype. We can get
rid of this warning by explicitly
defining the type of the values in
this column, which we can do by
using the dtype parameter.
4. To see the dtype of each column,

we can use the dtypes attribute of
a DataFrame:
# looking at the data

types of each column
Dataset.dtypes

Figure 5.18: Datatypes of each column

of the dataset
Note
Here, we can see the datatypes of

each column. Since the String
type is not a primitive datatype, it's
displayed as an object.
5. We use the head() method of

pandas DataFrames to display the
first five entries:
# showing the first 5

entries of the dataset
dataset.head()

Figure 5.19: First five entries of the
dataset
6. Most datasets won't be in the

format that you desire. Some of
them might have their Latitude
and Longitude values hidden in
a different column. This is where
the data wrangling skills of Chapter
1, The Importance of Data
Visualization and Data Exploration,
are needed. For the given dataset,
the transformations are easy – we
simply need to map the Latitude
and Longitude columns into the
lat and lon columns, which are
done by an assignment:
# mapping Latitude to
lat and Longitude to
lon
dataset['lat'] =
dataset['Latitude']
dataset['lon'] =
dataset['Longitude']
7. Our dataset is now ready for the

first plotting. We'll use a
DotDensityLayer here, which is a
good way to start and see all of our
data points:
# plotting the whole

dataset with dots
)
geoplotlib.show()

all the cities
8. Before we start breaking down our

data to get a better and more
workable dataset, we want to
understand the outlines of our full
data. We'll do this by displaying the
number of countries and the
number of cities that our dataset
holds:
# amount of countries
and cities
print(len(dataset.grou
pby(['Country'])),
'Countries')
print(len(dataset),
'Cities')

Figure 5.21: Grouping by countries and

cities
9. To see each grouped element on its

own, we use the size() method,
which returns a Series object:
# amount of cities per

country (first 20
entries)
dataset.groupby(['Coun
try']).size().head(20)

Figure 5.22: Number of cities per

country
10. We also want to display the average

amount of cities per country.
Aggregation is an important
concept in pandas that helps us in
achieving this:
# average num of
cities per country
dataset.groupby(['Coun
try']).size().agg('mea
n')

Figure 5.23: Average amount of cities

per country
11. We now want to reduce the amount

of data we are working with. One
approach would be to remove all
cities that don't have a population
value:
Note
Breaking down and filtering your

data is one of the most important
aspects of getting good insights.
Cluttered visualizations can hide
information.
# filter for countries
with a population
entry (Population > 0)
dataset_with_pop =
dataset[(dataset['Popu
lation'] > 0)]
print('Full dataset:',
len(dataset))
print('Cities with
population
information:',
len(dataset_with_pop))

Figure 5.24: Cities with population

information
12. Displaying the first five items from

the new dataset gives us a basic
indication of what values will be
present in the Population column:
# displaying the first

5 items from
dataset_with_pop
dataset_with_pop.head(
)

Figure 5.25: First five items of the
reduced dataset
13. Now. we can take a look at our

reduced dataset with the help of a
dot-densityplot:
# showing all cities

with a defined
population with a dot
density plot
_with_pop)
geoplotlib.show()
In the new dot plot, we can already

feel some improvements in terms of
clarity. We still have too many dots
on our map. Given the activity
definition, we can filter our dataset
further by only looking at cities
with a population of more than
100k.
14. To further filter our dataset, we can

use the same approach that we used
previously:
# dataset with cities

with population of >=
100k
dataset_100k =
dataset_with_pop[(data
set_with_pop['Populati
on'] >= 100_000)]
print('Cities with a
population of 100k or
more:',
len(dataset_100k))

Figure 5.26: Cities with a population of

100k or more
15. In addition to just plotting our 100k

dataset, we also want to fix our
viewport to a specific bounding
box. Since our data is spread across
the world, we can use the in-built
WORLD constant of the
BoundingBox class:
# displaying all
cities >= 100k
population with a
fixed bounding box
(WORLD) in a dot
density plot
import BoundingBox
_100k)
geoplotlib.set_bbox(Bo
undingBox.WORLD)
geoplotlib.show()

cities with a population of 100k or more
16. Comparing it with the previous

plots, it gives us a better view on
where large amounts of cities are
with a population of more than
100k. We now want to find the
areas of these cities that are the
most densely packed. As we've
seen in the previous exercise,
Voronoi is convenient for such
insights:
# using filled voronoi

to find dense areas
geoplotlib.voronoi(dat
aset_100k,
cmap='hot_r',
max_area=1e3,
alpha=255)
geoplotlib.show()
The resulting visualization is

exactly what we were searching for.
In the Voronoi plot, we can see
clear tendencies. Germany, Great
Britain, Nigeria, India, Japan, Java,
the east coast of the USA, and
Brazil stick out. We can now filter
our data and only look at those
countries to find the ones that are
best suited.
Note
You can also create custom color

map gradients with the ColorMap
class.
17. The last step is to reduce our

dataset to countries in Europe, such
as Germany and Great Britain. We
can use operators when supplying
conditions to filter our data. This
operator will allow us to filter for
Germany and Great Britain at the
same time:
# filter 100k dataset

for cities in Germany
and GB
dataset_europe =
dataset_100k[(dataset_
100k['Country'] ==
'de') |
(dataset_100k['Country
'] == 'gb')]
print('Cities in
Germany or GB with
population >= 100k:',
len(dataset_europe))
Figure 5.28: Cities in Germany and Great

Britain with a population of 100K or
more
18. Using Delaunay triangulation to

find the area with the most densely
packed cities, we can see one more
hotspot. The areas around Cologne,
Birmingham, and Manchester really
stick out:
#using Delaunay
triangulation to find
the most densely
populated area
geoplotlib.delaunay(da
taset_europe,
cmap='hot_r')
geoplotlib.show()
By using the hot_r color map, we

can quickly get a good visual
representation and make the areas
of interest pop out.

Geoplotlib. We made use of different plots to get the
information we needed. In the following activities, we will
look at some more custom features of Geoplotlib to change the
map tiles provider and create custom plotting layers.

WITH CUSTOM LAYERS
Solution:
Let's create a custom layer that will allow us to display

geospatial data and animate the data points over time:
1. In this first step, we only need

pandas for the data import:
# importing the
import pandas as pd
2. Use the read_csv method of

pandas to load the .csv file:

from the csv file
dataset =
pd.read_csv('./data/fl
ight_tracking.csv')
Note
The preceding dataset can be found

here: https://bit.ly/2DyPHwD.
3. We need to understand the structure

of our dataset by looking at the
provided features:

5 rows of the dataset
dataset.head()
Figure 5.29: First five elements of the
dataset
4. Remember that Geoplotlib needs

latitude and longitude
columns with the names lat and
lon. Use the rename method
provided in pandas to rename the
columns:
# renaming columns
latitude to lat and
longitude to lon
dataset =
dataset.rename(index=s
tr, columns=
{"latitude": "lat",
"longitude": "lon"})
5. Taking another look at the first five

elements of the dataset, we can now
observe that the names of the
columns have changed to lat and
lon:

dataset.head()
Figure 5.30: Dataset with the lat and lon

columns
6. Since we want to get a visualization

over time in this activity, we need
to work with date and time. If
we take a closer look at our dataset,
it shows us that date and time
are separated by two columns.
We need to combine date and time

into a timestamp using the
to_epoch method. The following
code does this:
# method to convert
date and time to an
unix timestamp
datetime
def to_epoch(date,
time):
try:
timestamp =
round(datetime.strptim
e('{} {}'.format(date,
time), '%Y/%m/%d
%H:%M:%S.%f').timestam
p())
return timestamp
except ValueError:
return
round(datetime.strptim
e('2017/09/11
17:02:06.418',
'%Y/%m/%d
%H:%M:%S.%f').timestam
p())
7. With the preceding method, we can

now use the apply method
provided by the pandas DataFrame
to create a new column called
timestamp that holds the unix
timestamp:
# creating a new
column called
timestamp with the
to_epoch method
applied
dataset['timestamp'] =
dataset.apply(lambda
x: to_epoch(x['date'],
x['time']), axis=1)
8. Taking another look at our dataset,

we now have a new column that
holds the unix timestamps:

dataset.head()
Figure 5.31: Dataset with a timestamp

column added
Since our dataset is now ready to be

used with all the necessary columns
in place, we can start writing our
custom layer. This layer will
display each point once it reaches
the timestamp that's provided in
the dataset. It will be displayed for
a few seconds before it disappears.
We'll need to keep track of the
current timestamp in our custom
layer. Consolidating what we
learned in the theoretical section of
this topic, we have a __init__
method that constructs our custom
TrackLayer.
9. In our draw method, we filter our

dataset for all the elements that are
in the mentioned time range and
use each element of the filtered list
to display it on the map with color
that's provided by the
colorbrewer method.
Since our dataset only contains data

from a specific time range and
we're always incrementing the time,
we want to check whether there are
still any elements with
timestamps after the current
timestamp. If not, we want to set
our current timestamp to the earliest
timestamp that's available in the
dataset. The following code shows
how we can create a custom layer:
# custom layer
creation
import geoplotlib
from geoplotlib.layers
import BaseLayer
from geoplotlib.core
import BatchPainter
from geoplotlib.colors
import colorbrewer
import epoch_to_str,
BoundingBox
class
TrackLayer(BaseLayer):
def __init__(self,
dataset,
bbox=BoundingBox.WORLD
):
self.data = dataset
self.cmap =
colorbrewer(self.data[
'hex_ident'],
alpha=200)
self.time =
self.data['timestamp']
.min()
self.painter =
BatchPainter()
self.view = bbox
def draw(self, proj,

mouse_x, mouse_y,
ui_manager):
self.painter =
BatchPainter()
df =
self.data.where((self.
data['timestamp'] >
self.time) &
(self.data['timestamp'
] <= self.time + 180))
for element in
set(df['hex_ident']):
grp =
df.where(df['hex_ident
'] == element)
self.painter.set_color
(self.cmap[element])
x, y =
proj.lonlat_to_screen(
grp['lon'],
grp['lat'])
self.painter.points(x,
y, 15, rounded=True)
self.time += 1
if self.time >
.max():
self.time =
.min()
self.painter.batch_dra
w()
ui_manager.info('Curre
nt timestamp:
{}'.format(epoch_to_st
r(self.time)))
# bounding box that

gets used when layer
is created
def bbox(self):
return self.view
10. Since our dataset only contains data

from the area around Leeds in the
UK, we need to define a custom
BoundingBox that focuses our
view on this area:
# bounding box for our

view on Leeds
import BoundingBox
leeds_bbox =
BoundingBox(north=53.8
074, west=-3,
south=53.7074 ,
east=0)
11. Geoplotlib sometimes requires you

to provide a
DataAccessObject instead of a
pandas DataFrame. Geoplotlib
provides a convenient method for
converting any pandas DataFrame
into a DataAccessObject:
# displaying our
custom layer using
add_layer
import
DataAccessObject
data =
DataAccessObject(datas
et)
geoplotlib.add_layer(T
rackLayer(data,
bbox=leeds_bbox))
geoplotlib.show()
Congratulations! You've completed the custom layer activity
using Geoplotlib. We've applied several pre-processing steps to
shape the dataset as we want to have it. We've also written a
custom layer to display spatial data in the temporal space. Our
custom layer even has a level of animation. This is something
we'll look into more in the following chapter about Bokeh.
Chapter 6: Making
Things Interactive
with Bokeh
PLOTS WITH WIDGETS
Solution:
1. Open the
this activity.
# importing the
import pandas as pd
3. Again, we want to display our plots

inside a Jupyter Notebook, so we
have to import and call the
output_notebook method from
the io interface of Bokeh:

figures inside the
notebook

output_notebook
output_notebook()

olympia2016_athletes.csv
data:

with geoplotlib
dataset =
pd.read_csv('./data/ol
ympia2016_athletes.csv
')

# looking at the
dataset
dataset.head()

Figure 6.25: Loading the top five rows of the
olympia2016_athletes dataset using the head method
Building an Interactive Visualization
1. To create our visualization, we need

some additional imports. We'll
again import figure and show
from the plotting interface, which
will give us the tools we need to
create our plot. The widgets, as we
saw in the previous exercise, come
from the ipywidgets library.
Here, we'll be using interact,
again, as decorator, and the
widgets interface, which gives us
access to different widgets:
# importing the
from bokeh.plotting
import figure, show

interact, widgets
2. Like in the previous exercises, we

need to do some data extraction
first. In this activity, we will need a
list of unique countries from the
dataset, the amount of athletes for
each country, and the amount of
medals won by each country, split
into gold, silver, and bronze:
# extract countries
and group Olympians by
country and their sex
# and the number of

medals per country by
sex
countries =
dataset['nationality']
.unique()
athletes_per_country =
dataset.groupby('natio
nality').size()
medals_per_country =
dataset.groupby('natio
nality')['gold',
'silver','bronze'].sum
()
3. Before we go in and implement the

plotting for this visualization, we
want to set up our widgets and
the @interact method that will
later display the plot upon
execution. Execute this
empty get_plot() method cell
and then move on to the widget
creation. We will implement this
later.
4. In this activity, we will use

two IntSlider widgets that will
control the max numbers for the
amount of athletes and/or medals a
country is allowed to have in order
to be displayed in the
visualization. We need two values
to set up the widgets: the maximum
amount of medals of all the
countries and the maximum amount
of athletes of all the countries:
# getting the max

amount of medals and
athletes of all
countries
max_medals =
medals_per_country.sum
(axis=1).max()
max_athletes =
athletes_per_country.m
ax()
5. Using those maximum numbers as

the maximum for both widgets will
give us reasonable slider values that
are dynamically adjusted if we
should increase the amount of
athletes or medals in the dataset.
We need two IntSlider objects
that handle the input for
our max_athletes and max_me
dals. To look like our actual
visualization, we want to have
the max_athletes_slider dis
played in a vertical orientation and
the max_medals_slider in a
horizontal orientation.In the
visualization, they should be
displayed as Max. Athletes
and Max. Medals:
# setting up the
interaction elements
max_athletes_slider=wi
dgets.IntSlider(value=
max_athletes, min=0,
max=max_athletes,
step=1,
description='Max.
Athletes:',
e,
orientation='vertical'
, layout={'width':
'100px'})
max_medals_slider=widg
ets.IntSlider(value=ma
x_medals, min=0,
max=max_medals,
step=1,
description='Max.
Medals:',
e,
orientation='horizonta
l')
6. After setting up the widgets, we can

implement the method that will be
called with each update of the
interaction widgets. As we saw in
the previous exercise, we will use
the @interact decorator for this.
Instead of value ranges or lists, we
will provide the variable names of
our already created widgets in the
decorator. Since we have already
set up the empty method that will
return a plot, we can call show()
with the method call inside it to
show the result once it is returned
from the get_plot method:
# creating the
interact method
@interact(max_athletes
=max_athletes_slider,
max_medals=max_medals_
slider)
def
get_olympia_stats(max_
athletes, max_medals):
show(get_plot(max_athl
etes, max_medals))
7. As we mentioned in the previous

exercise, we can also make use of
the widgets that are described here:
https://ipywidgets.readthedocs.io/en
/stable/examples/Widget%20List.ht
ml.
8. Once you've built the widgets, upon

execution, you will see them being
displayed below the cell. We are
now ready to scroll up and
implement the plotting with Bokeh.
9. The two arguments we get passed

are max_athletes and max_me
dals. Both of them are int
values. First, we want to filter down
our countries dataset that contains
all the countries that placed athletes
in the Olympic Games. We need to
check whether they have less than
or equal medals and athletes than
our max values, passed as
arguments. Once we have a filtered
down dataset, we can create our
DataSource. This DataSource will
be used both for the tooltips and the
printing of the circle glyphs.
Note
There is extensive documentation

on how to use and set up tooltips
that you can and use, which can be
accessed from the following
link: https://bokeh.pydata.org/en/lat
est/docs/user_guide/tools.html.
10. Create a new plot using

the figure method that has the
following attributes: title of 'Rio
comparison', x_axis_label
of 'Number of Medals', and
y_axis_label of 'Num of
Athletes':
# creating the scatter

plot
def
get_plot(max_athletes,
max_medals):
filtered_countries=[]
for country in
countries:
if
(athletes_per_country[
country] <=
max_athletes and
medals_per_country.loc
[country].sum() <=
max_medals):
filtered_countries.app
end(country)
data_source=get_dataso
urce(filtered_countrie
s)
TOOLTIPS=[ ('Country',
'@countries'),('Num of
Athletes', '@y'),
('Gold', '@gold'),
('Silver', '@silver'),
('Bronze', '@bronze')]
plot=figure(title='Rio
comparison',
x_axis_label='Number
of Medals',
y_axis_label='Num of
Athletes',
plot_width=800,
plot_height=500,
tooltips=TOOLTIPS)
plot.circle('x', 'y',
source=data_source,
size=20,
color='color',
alpha=0.5)
return plot
11. To display every country with a

different color, we want to
randomly create the colors with a
six-digit hex code. The following
method does this:
# get a 6 digit random

hex color to
differentiate the
countries better
import random
def
get_random_color():
return '%06x' %
random.randint(0,
0xFFFFFF)
12. We will use a Bokeh

ColumnDataSource to handle
our data and make it easily
accessible for our tooltip and
glyphs. Since we want to display
additional information in a tooltip,
we need our DataSource to have the
color field, which holds the
required amount of random colors;
the countries field, which holds
the list of filtered down countries;
the gold, silver, and bronze
fields, which hold the number of
gold, silver, and bronze
medals for each country,
respectively; the x field, which
holds the summed number of
medals for each country; and the
y field, which holds the number of
athletes for each country:
# build the DataSource
def
get_datasource(filtere
d_countries):
return
ColumnDataSource(data=
dict(
color=
[get_random_color()
for _ in
filtered_countries],
countries=filtered_cou
ntries,
gold=
[medals_per_country.lo
c[country]['gold'] for
country in
silver=
c[country]['silver']
for country in
bronze=
c[country]['bronze']
for country in
x=
c[country].sum() for
country in
y=
[athletes_per_country.
loc[country].sum() for
country in
filtered_countries]
))
13. After our implementation is

finished, we can execute the last
cell with our @interact
decorator once more. This time, it
will display our scatter plot with
our interactive widgets. We will see
each country in a different color.
Upon hovering over them, we will
get more information about each
country, such as its short name,
number of athletes, and the amount
of gold, silver, and bronze medals
they earned. The resulting
visualization should look as
follows:
Figure 6.26: Final interactive visualization that
displays the scatter plot
You've built a full visualization to display and explore data

from the 2016 Olympics. We added two widgets to our
visualization that allows us to filter down our displayed
countries. As we mentioned before, when working with
interactive features and Bokeh, you might want to read up
about Bokeh Server a little bit more. It will give you more
options to express your creativity by creating animated plots
and visualizations that can be explored by several people at the
same time.
Chapter 7: Combining
What We Have
Learned
ACTIVITY 30:
IMPLEMENTING
MATPLOTLIB AND
SEABORN ON NEW YORK
CITY DATABASE
Solution:

b from the Lesson07 folder to
implement this activity. Import all
the necessary libraries:
# Import statements
import pandas as pd
import numpy as np
import matplotlib
import
plt
import squarify
sns.set()
2. Use pandas to read both .csv files

located in the data subdirectory:
p_ny =
pd.read_csv('./data/pn
y.csv')
h_ny =
pd.read_csv('./data/hn
y.csv')
3. Use the given PUMA (public use

microdata area code based on the
2010 Census definition, which are
areas with populations of 100,000
or more) ranges to further divide
the dataset into NYC districts
(Bronx, Manhatten, Staten Island,
Brooklyn, and Queens):
# PUMA ranges
bronx = [3701, 3710]
manhatten = [3801,
3810]
staten_island = [3901,
3903]
brooklyn = [4001,
4017]
queens = [4101, 4114]
nyc = [bronx[0],
queens[1]]
def puma_filter(data,
puma_ranges):
return
data.loc[(data['PUMA']
>= puma_ranges[0]) &
(data['PUMA'] <=
puma_ranges[1])]
h_bronx =
puma_filter(h_ny,
bronx)
h_manhatten =
puma_filter(h_ny,
manhatten)
h_staten_island =
puma_filter(h_ny,
staten_island)
h_brooklyn =
puma_filter(h_ny,
brooklyn)
h_queens =
puma_filter(h_ny,
queens)
p_nyc =
puma_filter(p_ny, nyc)
h_nyc =
puma_filter(h_ny, nyc)
4. In the dataset, each sample has a

certain weight that reflects the
weight for the total dataset.
Therefore, we cannot simply
calculate the median. Use the given
weighted_median function in
the following code to compute the
median:
# Function for a
'weighted' median
def
weighted_frequency(val
ues, weights):
weighted_values = []
for value, weight in

zip(values, weights):
weighted_values.ex
tend(np.repeat(value,
weight))
return
weighted_values
def
weighted_median(values
, weights):
return
np.median(weighted_fre
quency(values,
weights))
5. In this subtask, we will create a plot

containing multiple subplots that
visualize information with regards
to NYC wages. Visualize the
median household income for the
US, New York, New York City, and
its districts. Visualize the average
wage by gender for the given
occupation categories for the
population of NYC. Visualize the
wage distribution for New York and
NYC. Use the following yearly
wage intervals: 10k steps between 0
and 100k, 50k steps between 100k
and 200k, and >200k:
Lesson07/Activity30/activity30_solution.
ipynb
# Data wrangling for

median housing income
income_adjustement =
h_ny.loc[0,
['ADJINC']].values[0]
/ 1e6
def
median_housing_income(
data):
//[…]
h_queens_income_median
=
median_housing_income(
h_queens)

wage by gender for
different occupation
categories
occ_categories =
['Management,\nBusines
s,\nScience,\nand
Arts\nOccupations',
'Service\nOccupations'
,
'Sales
and\nOffice\nOccupatio
ns', 'Natural
Resources,\nConstructi
on,\nand
Maintenance\nOccupatio
ns',
'Production,\nTranspor
tation,\nand Material
Moving\nOccupations']
//[…]
wages_female =
wage_by_gender_and_occ
upation(p_nyc, 2)

wage distribution
wage_bins = {'<$10k':
[0, 10000], '$10-20k':
[10000, 20000], '$20-
30k': [20000, 30000],
'$30-40k': [30000,
40000], '$10-20k':
[40000, 50000],
'$50-60k': [50000,
60000], '$60-70k':
[60000, 70000], '$70-
80k': [70000, 80000],
'$80-90k': [80000,
90000], '$90-100k':
[90000, 100000],
'$100-150k': [100000,
150000], '$150-200k':
[150000, 200000],
'>$200k': [200000,
np.infty]}
//[…]
wages_ny =
wage_frequency(p_ny)
# Create figure with

four subplots
fig, (ax1, ax2, ax3) =

plt.subplots(3, 1,
figsize=(7, 10),
dpi=300)
# Median household
income in the US
us_income_median =
60336
# Median household
income
ax1.set_title('Median
Household Income',
fontsize=14)
//[…]
ax1.set_xlabel('Yearly
household income in
$')
# Wage by gender in
common jobs
ax2.set_title('Wage by
Gender for different
Job Categories',
fontsize=14)
x = np.arange(5) + 1
//[…]
ax2.set_ylabel('Averag
e Salary in $')
# Wage distribution
ax3.set_title('Wage
Distribution',
fontsize=14)
x =
np.arange(len(wages_ny
c)) + 1
width = 0.4
//[…]
ax3.vlines(x=9.5,
ymin=0, ymax=15,
linestyle='--')
# Overall figure
fig.tight_layout()
plt.show()
https://bit.ly/2StchfL

Figure 7.15: Wage statistics for New
York City in comparison with New York
and the United States

percentage for the given occupation
subcategories for the population of
NYC:
ipynb

occupations
occ_subcategories =
{'Management,\nBusines
s,\nand Financial':
[10, 950],
//[..]
def
occupation_percentage(
data):
percentages = []
overall_sum =
np.sum(data.loc[(data[
'OCCP'] >= 10) &
(data['OCCP'] <=
9750),
['PWGTP']].values)
for occ in
occ_subcategories.valu
es():
query =
data.loc[(data['OCCP']
>= occ[0]) &
(data['OCCP'] <=
occ[1]),
['PWGTP']].values
percentages.append(np.
sum(query) /
overall_sum)
return percentages
occ_percentages =
occupation_percentage(
p_nyc)
# Visualization of
tree map
plt.figure(figsize=
(16, 6), dpi=300)
//[..]
plt.axis('off')
plt.title('Occupations
in New York City',
fontsize=24)
plt.show()
https://bit.ly/2StchfL

Figure 7.16: Occupations in NYC
7. Use a heatmap to show the

correlation between difficulties
(self-care difficulty, hearing
difficulty, vision, difficulty,
independent living difficulty,
ambulatory difficulty, veteran
service-connected disability, and
cognitive difficulty) and age groups
(<5, 5-11, 12-14, 15-17, 18-24, 25-
34, 35-44, 45-54, 55-64, 65-74, and
75+) in New York City:

New York City
population
difficulties
difficulties = {'Self-
care difficulty':
'DDRS', 'Hearing
difficulty': 'DEAR',
'Vision difficulty':
'DEYE', 'Independent
living difficulty':
'DOUT',
'Ambulatory
difficulty': 'DPHY',
'Veteran service
connected disability':
'DRATX',
'Cognitive
difficulty': 'DREM'}
age_groups = {'<5':
[0, 4], '5-11': [5,
11], '12-14': [12,
14], '15-17': [15,
17], '18-24': [18,
24], '25-34': [25,
34],
'35-44': [35, 44],

'45-54': [45, 54],
'55-64': [55, 64],
'65-74': [65, 74],
'75+': [75, np.infty]}
def
difficulty_age_array(d
ata):
array =
np.zeros((len(difficul
ties.values()),
len(age_groups.values(
))))
for d, diff in
enumerate(difficulties
.values()):
for a, age in
enumerate(age_groups.v
alues()):
age_sum =
np.sum(data.loc[(data[
'AGEP'] >= age[0]) &
(data['AGEP'] <=
age[1]),
['PWGTP']].values)
query =
data.loc[(data['AGEP']
>= age[0]) &
(data['AGEP'] <=
age[1]) & (data[diff]
== 1),
['PWGTP']].values
array[d, a] =
np.sum(query) /
age_sum
return array
array =
difficulty_age_array(p
_nyc)
# Heatmap
plt.figure(dpi=300)
ax = sns.heatmap(array
* 100)
ax.set_yticklabels(dif
ficulties.keys(),
rotation=0)
ax.set_xticklabels(age
_groups.keys(),
rotation=90)
ax.set_xlabel('Age
Groups')
ax.set_title('Percenta
ge of NYC population
with difficulties',
fontsize=14)
plt.show()

Figure 7.17: Percentage of NYC population with

difficulties
ACTIVITY 31: BOKEH

STOCK PRICES
VISUALIZATION
Solution:
1. Open the
this activity. Import the pandas
library:
# importing the
import pandas as pd
2. Display our plots inside a Jupyter

Notebook. We have to import and
call the output_notebook
Bokeh:

figures inside the
notebook

output_notebook
output_notebook()

stock_prices.csv data:

with geoplotlib
dataset =
pd.read_csv('./data/st
ock_prices.csv')

# looking at the
dataset
dataset.head()

Figure 7.18: Head of our imported data
5. Since the date column has no

information about the hour, minute,
and second, we want to avoid
displaying them in the visualization
later on and simply display the year,
month, and day. Therefore, we'll
create a new column that holds the
formatted short version of the date
value. The execution of the cell will
take a moment since it's a fairly
large dataset. Please be patient:
# mapping the date of

each row to only the
year-month-day format

datetime
def
shorten_time_stamp(tim
estamp):
shortened =
timestamp[0]
if len(shortened) >
10:
parsed_date=dateti
me.strptime(shortened,
'%Y-%m-%d %H:%M:%S')
shortened=datetime
.strftime(parsed_date,
'%Y-%m-%d')
return shortened
dataset['short_date']
= dataset.apply(lambda
x:
shorten_time_stamp(x),
axis=1)
6. Taking another look at our updated

dataset, we can see a new column
called short_date that holds the
date without the hour, minute, and
second information:
# looking at the
dataset with shortened
date
dataset.head()

Figure 7.19: Dataset with added short_date column
Building an Interactive Visualization
1. To create our visualization, we need

some additional imports. We'll
again import figure and show
this from the plotting interface,
which will give us the tools we
need to create our plot. The
widgets, as we saw in the previous
exercise, come from the
ipywidgets library. Here, we'll
be using @interact, again, as a
decorator, and the widgets
interface, which gives us access to
the different widgets:
# importing the
from bokeh.plotting
import figure, show

interact, widgets
2. Before we go in and implement the

plotting methods, we want to set up
the interactivity widgets. Scroll
down to the cell that says
#extracing the necessary
data before implementing the
plotting. Still make sure that you
execute the cells below that, even
though this will simply pass and do
nothing for now. We want to start
implementing our visualization
here. In the following cells, we will
extract the necessary data that will
be provided to the widget elements.
In the first cell, we want to extract
the following information: a list of
unique stock names that are present
in the dataset, a list of all
short_dates that are in 2016, a
sorted list of unique dates generated
from the previous list of dates from
2016, and a list with the values
open-close and volume.
3. Once we have this information in

place, we can start building our
widgets:
# extracing the
necessary data
stock_names=dataset['s
ymbol'].unique()
dates_2016=dataset[dat
aset['short_date'] >=
'2016-01-01']
['short_date']
unique_dates_2016=sort
ed(dates_2016.unique()
)
value_options=['open-
close', 'volume']
4. Given the extracted information

from the preceding cell, we can
now define widgets and provide
the available options for it. As we
mentioned in the introduction, we
want to have several interactive
features, including two dropdowns
with which we can select two
stocks that should be compared to
each other. The first dropdown by
default should have the AAPL stock
selected, named Compare:, while
the second dropdown by default
should have the AON stock
selected, named to::
# setting up the
interaction elements
drp_1=widgets.Dropdown
(options=stock_names,
value='AAPL',
description='Compare:'
)
drp_2=widgets.Dropdown
(options=stock_names,
value='AON',
description='to:')
5. Then, we need
SelectionRange, which will
allow us to select a range of dates
from the extracted list of unique
2016 dates. By default, the first 25
dates should be selected, named
From-To. Make sure to disable the
continuous_update
parameter. Adjust the layout width
to 500px to make sure that the
dates are displayed correctly:
range_slider=widgets.S
electionRangeSlider(op
tions=unique_dates_201
6,
index=(0,25),
e,
description='From-To',
layout={'width':
'500px'})
6. Add a RadioButton group that

provides the open-close and
volume options. By default,
open-close should be selected,
named Metric:
range_slider=widgets.S
electionRangeSlider(op
tions=unique_dates_201
6,
index=(0,25),
e,
description='From-To',
layout={'width':
'500px'})
value_radio=widgets.Ra
dioButtons(options=val
ue_options,
value='open-close',
description='Metric')
Note
As we mentioned in the previous

exercise, we can also make use of
the widgets that are described here:
https://bit.ly/2Te9jAf.
7. After setting up the widgets, we can

implement the method that will be
called with each update of the
interaction widgets. As we saw in
the previous exercise, we will use
the @interact decorator for this.
8. Instead of value ranges or lists, we

will provide the variable names of
our already created widgets in the
decorator. The method will get four
arguments: stock_1, stock_2,
date, and value. Since we have
already set up the empty method
that will return the preceding plot,
we can call show() with the
method call inside to show the
result once it is returned from the
get_stock_for_2016 method.
9. Once you've built the widgets, upon

execution, you will see them being
displayed below the cell:
# creating the
interact method
@interact(stock_1=drp_
1, stock_2=drp_2,
date=range_slider,
value=value_radio)
def
get_stock_for_2016(sto
ck_1, stock_2, date,
value):
show(get_plot(stock_1,
stock_2, date, value))
10. We are now ready to scroll up and

implement the plotting with Bokeh.
At the moment, our show() in the
last cell will not render any
elements into our visualization. We
will start with the so-called candle
stick visualization, which is often
used with stock price data.
11. The already defined method gets

our plot object, a stock_name,
a stock_range containing the
data of only the selected date range
that was defined with the widgets,
and a color for the line. We will use
those arguments to create the
candle sticks. They basically
contain a segment that creates the
vertical line, and either a green or
red vbar to color code whether the
close price is lower than the open
price. Once you have created the
candles, we also want to have a
continuous line running through the
mean (high, low) point of each
candle. So, you have to calculate
the mean for every (high/low) pair
and then plot those data points with
a line with the given color:
def
add_candle_plot(plot,
stock_name,
stock_range, color):
inc_1 =
stock_range.close >
stock_range.open
dec_1 =
stock_range.open >
stock_range.close
w = 0.5
plot.segment(stock_ran
ge['short_date'],
stock_range['high'],
stock_range['short_dat
e'],
stock_range['low'],
color="grey")
plot.vbar(stock_range[
'short_date'][inc_1],
w,
stock_range['high']
[inc_1],
stock_range['close']
[inc_1],
fill_color="green",
line_color="black",
legend=('Mean price of
' + stock_name),
muted_alpha=0.2)
plot.vbar(stock_range[
'short_date'][dec_1],
w,
stock_range['high']
[dec_1],
stock_range['close']
[dec_1],
fill_color="red",
line_color="black",
' + stock_name),
muted_alpha=0.2)
stock_mean_val=stock_r
ange[['high',
'low']].mean(axis=1)
plot.line(stock_range[
'short_date'],
stock_mean_val,
' + stock_name),
muted_alpha=0.2,
line_color=color,
alpha=0.5)
Note
Make sure to reference the example

provided in the Bokeh library here.
You can adapt the code in there to
our arguments:
ocs/gallery/candlestick.html.
12. After you are done implementing

the add_candle_plot method,
scroll down and run the
@interact cell again. You will
now see the candles being
displayed for the two selected
stocks. The last missing step is
implementing the plotting of the
lines if the volume value is
selected.
13. One additional interaction feature is

to have an interactive legend that
allows us to mute, meaning gray
out, each stock in the visualization:
Lesson07/Activity31/activity31_solution.ipynb
# method to build the plot
def get_plot(stock_1, stock_2, date,

value):
//[..]
plot.xaxis.major_label_orientation = 1
plot.grid.grid_line_alpha=0.3
if value == 'open-close':
add_candle_plot(plot, stock_1_name,
stock_1_range, 'blue')
add_candle_plot(plot, stock_2_name,
stock_2_range, 'orange')
if value == 'volume':
plot.line(stock_1_range['short_date'],
stock_1_range['volume'],
legend=stock_1_name,
muted_alpha=0.2)
plot.line(stock_2_range['short_date'],
stock_2_range['volume'],
legend=stock_2_name,
muted_alpha=0.2,
line_color='orange')
plot.legend.click_policy="mute"
return plot
https://bit.ly/2GRneWR
Note
To make our legend interactive please take a look at the
documentation for the legend feature:
https://bokeh.pydata.org/en/latest/docs/user_guide/interaction/
legends.html.
After our implementation has finished, we can execute the last

cell with our @interact decorator once more. This time, it
will display our candlestick plot and once we switch to the
volume RadioButton, we will see the volumes displayed that
have been traded at the given dates. The resulting visualization
should look somewhat like this:
displays the candlestick plot
The following figure shows the final interactive visualization

of volume plot:
displays the volume plot
Congratulations!
You've built a full visualization to display and explore stock

price data. We added several widgets to our visualization that
allows us to select to be compared stocks, restrict the displayed
data to a specific date range, and even display two different
kind of plots.
As we mentioned before, when working with interactive

features and Bokeh, you might want to read up about Bokeh
Server a little bit more. It will give you more options to express
your creativity by creating animated plots and visualizations
that can be explored by several people at the same time.

AIRBNB DATA WITH
GEOPLOTLIB
Solution:
1. Open the
activity032_solution.ipy
nb Jupyter Notebook from the
this activity. Import NumPy,
pandas, and Geoplotlib first:
# importing the
import numpy as np
import pandas as pd
import geoplotlib
2. Use the read_csv method of

pandas to load the .csv file. If
your computer is a little slow, use
the smaller dataset:
dataset =
pd.read_csv('./data/ai
rbnb_new_york.csv')
# dataset =
pd.read_csv('./data/ai
rbnb_new_york_smaller.
csv')
3. Understand the structure of our

dataset by looking at the provided
features:
# print the first 5

rows of the dataset
dataset.head()

Figure 7.22: Displaying the first five
elements of the dataset
4. Remember that Geoplotlib needs

latitude and longitude
columns with the names lat and
lon. We will therefore add new
columns for lat and lon and
assign the corresponding value
columns to it:
# mapping Latitude to
lat and Longitude to
lon
dataset['lat'] =
dataset['latitude']
dataset['lon'] =
dataset['longitude']
5. When creating a color map that

changes color based on the price of
an accommodation, we need a
value that can easily be compared
and checked whether it's smaller or
bigger than any other listing.
Therefore, we will create a new
column called dollar_price
that will hold the value of the price
column as a float:
# convert string of
type $<numbers> to
<nubmers> of type
float
def
convert_to_float(x):
try:
value=str.replace(x[1:
], ',', '')
return float(value)
except:
return 0.0
# create new
dollar_price column
with the price as a
number
# and replace the NaN

values by 0 in the
ratings column
dataset['price'] =
dataset['price'].filln
a('$0.0')
dataset['review_scores
_rating'] =
dataset['review_scores
_rating'].fillna(0.0)
dataset['dollar_price'
] =
dataset['price'].apply
(lambda x:
convert_to_float(x))
6. This dataset has 96 columns. When

working with such a huge dataset, it
makes sense to think about what
data we really need and create a
subsection of our dataset that only
holds the data we need. Before we
can do that, we'll take a look at all
the columns that are available and
an example for that column. This
will help us decide what
information is suitable:
# print the col name
and the first entry
per column
for col in
dataset.columns:
print('{}\t{}'.format(
col, dataset[col][0]))

Figure 7.23: Each column header with
an example entry from the dataset
7. For now, we want to only use the

fields that help us build the
described visualization. Those
fields are id, latitude (as
lat), longitude (as lon),
price (in $), and
review_scores_rating:
# create a subsection
of the dataset with
the above mentioned
columns
columns=['id', 'lat',
'lon', 'dollar_price',
'review_scores_rating'
]
sub_data=dataset[colum
ns]
8. Taking another look at our dataset,

we now have a new column that
holds the unix timestamps:
# print the first 5

rows of the dataset
sub_data.head()

Figure 7.24: Displaying the first five
rows after keeping only five columns
9. Even though we know that our data

holds Airbnb listings for New York
City, at the moment, we have no
feeling about the amount,
distribution, and character of our
dataset. The simplest way to get a
first glance at the data is to plot
every listing with a simple dot map:
# import
DataAccessObject and
create a data object
as an instance of that
class
import
DataAccessObject
data =
DataAccessObject(sub_d
ata)
# plotting the whole

dataset with dots
geoplotlib.dot(data)
geoplotlib.show()

Figure 7.25: Simple dot map created
from the points
10. The last step is to write the custom

layer. Here, we want to define a
ValueLayer that extends the
BaseLayer of Geoplotlib. For the
mentioned interactive feature, we
need an additional import. pyglet
provides us with the option to act
on key presses. Given the data, we
want to plot each point on the map
with a color that is defined by the
currently selected attribute, either
price or rating.
11. To avoid non-descriptive output, we

need to also adjust the scale of our
color map. Ratings are between 0
and 100, whereas prices can be
much higher. Using a linear (lin)
scale for the ratings and a
logarithmic (log) scale for the
price will give us much better
insights into our data.
12. The view (bounding box) of our

visualization will be set to New
York and text information with the
currently selected attribute will be
displayed in the upper right corner:
Figure 7.26: jet color map scale
13. To assign each point a different

color, we simply paint each point
separately. This is definitely not the
most efficient solution, but it will
do for now. We will need the
following instance variables:
self.data, which holds the
dataset, self.display, which
holds the currently selected
attribute name, self.painter,
which holds an instance of the
BatchPainter class,
self.view, which holds the
BoundingBox, self.cmap,
which holds a color map with the
jet color schema, and an alpha of
255 and 100 levels.
14. Inside the invalidate method,

which holds the logic of projection
of the data to points on the map, we
have to switch between the lin
and log scales, depending on the
attribute that is currently selected.
The color is then determined by
placing the value between 0/1 and
the maximum (max_val) value,
which also has to be taken from the
dataset based on what attribute is
currently being displayed:
ipynb
# custom layer
creation
import pyglet
import geoplotlib
//[..]
class
ValueLayer(BaseLayer):
def __init__(self,
dataset,
bbox=BoundingBox.WORLD
):
//[..]
def invalidate(self,
proj):
# paint every point

with a color that
represents the
currently selected
attributes value
self.painter =
BatchPainter()
max_val =
max(self.data[self.dis
play])
scale = 'log' if
self.display ==
'dollar_price' else
'lin'
for index, id in
enumerate(self.data['i
d']):
//[..]
def draw(self, proj,

mouse_x, mouse_y,
ui_manager):
# display the ui
manager info
ui_manager.info('Use
left and right to
switch between the
displaying of price
and ratings. Currently
displaying:
{}'.format(self.displa
y))
self.painter.batch_dra
w()
def
on_key_release(self,
key, modifiers):
//[..]
# bounding box that

gets used when layer
is created
def bbox(self):
return self.view
https://bit.ly/2VoQveT
15. Since our dataset only contains data

from New York, we want to set the
view to New York in the beginning.
Therefore, we need an instance of
the BoundingBox class with the
given parameters. In addition to a
custom BoundingBox, we will
use the darkmatter tile provider
that we looked at in Chapter 5,
Plotting Geospatial Data:
# bounding box for our

view on New York
import BoundingBox
ny_bbox =
BoundingBox(north=40.8
97994,
west=-73.999040,
south=40.595581,
east=-73.95040)
# displaying our
custom layer using
add_layer
der('darkmatter')
geoplotlib.add_layer(V
alueLayer(data,
bbox=ny_bbox))
geoplotlib.show()
16. After launching our visualization,

we can see that our viewport is
focused on New York. Every
accommodation is displayed with
one dot that is colored based on
either its price, or upon clicking the
right or left arrow, the rating. We
can see that the general color gets
closer to yellow/orange the closer
we get to central Manhattan. On the
other hand, in the ratings
visualization, we can see that the
accommodations in central
Manhattan seem to be rated lower
than the ones outside:
Figure 7.27: New York Airbnb dot map, colored based
on the price
The following figure shows a dot map with color based on

rating:
Figure7.28: New York Airbnb dot map, colored based
on the ratings
Congratulations!
You've created an interactive visualization by writing your own

custom layer to display and visualize price and rating
information for Airbnb accommodations spread across New
York. As we can now see, writing custom layers for Geoplotlib
is a good approach for focusing on the attributes that you are
interested in.
Thanks for picking up this course to enhance your data

visualization skills with Python.

Data Visualization With Python PDF

Uploaded by

Data Visualization With Python PDF

Uploaded by

Contents

1. About the Book

1. About the Author

1. Introduction to Data Visualization

1. Measures of Central Tendency

1. Exercise 1: Loading a Sample Dataset and

1. Advantages of pandas over NumPy

7. What Makes a Good Visualization?

1. Activity 11: Identifying the Ideal Visualization

4. Basic Text and Legend Functions

1. Basic Image Operations

8. Writing Mathematical Expressions

1. Seaborn Figure Styles

1. Categorical Color Palettes

4. Interesting Plots in Seaborn

1. Activity 25: Linear Regression

1. Activity 26: Water Usage Revisited

1. The Design Principles of Geoplotlib

1. Exercise 8: Visually Comparing Different Tile

1. Activity 28: Working with Custom Layers

1. Exercise 11: Basic Interactivity Widgets

1. Activity 30: Implementing Matplotlib and

1. Chapter 1: The Importance of Data Visualization and Data

1. Activity 1: Using NumPy to Compute the Mean,

2. Chapter 2: All You Need to Know about Plots

1. Activity 7: Employee Skill Comparison

3. Chapter 3: A Deep Dive into Matplotlib

1. Activity 12: Visualizing Stock Trends by Using a

4. Chapter 4: Simplifying Visualizations Using Seaborn

1. Activity 20: Comparing IQ Scores for Different

5. Chapter 5: Plotting Geospatial Data

1. Activity 27: Plotting Geospatial Data on a Map

6. Chapter 6: Making Things Interactive with Bokeh

1. Activity 29: Extending Plots with Widgets

7. Chapter 7: Combining What We Have Learned

1. Activity 30: Implementing Matplotlib and

All rights reserved. No part of this book may be reproduced,

Every effort has been made in the preparation of this book to

Packt Publishing has endeavored to provide trademark

Author: Mario Döbler and Tim Größmann

Reviewer: José España

Managing Editor: Aditya Shah and Bhavesh Bangera

Acquisitions Editor: Kunal Sawant

Production Editor: Shantanu Zagade

Editorial Board: David Barnes, Ewan Buckingham, Shivangi

Production Reference: 1280219

All You Need to Know

STACKED BAR CHART

WHAT MAKES A GOOD

A Deep Dive into

ACTIVITY 12: VISUALIZING

ACTIVITY 13: CREATING A

ACTIVITY 14: CREATING A

ACTIVITY 15: COMPARING

ACTIVITY 17: USING A

ACTIVITY 18: CREATING

ACTIVITY 19: PLOTTING

REMOVING AXES SPINES

ACTIVITY 20: COMPARING

ACTIVITY 21: USING

ACTIVITY 22: MOVIE

ACTIVITY 23: COMPARING

ACTIVITY 24: TOP 30

ACTIVITY 25: LINEAR

ACTIVITY 26: WATER

ACTIVITY 27: PLOTTING

ACTIVITY 28: WORKING

EXERCISE 11: BASIC

ACTIVITY 29: EXTENDING

ACTIVITY 31: VISUALIZING

ACTIVITY 32: ANALYZING

About the Book

After you get a hang of the various visualization libraries,

This book ends with an interesting activity in which you will

ABOUT THE AUTHOR

Tim Großmann is a CS student with an interest in diverse