Advanced Visualizations & Dashboards:

Visualization Using Python

Explore approaches to building and implementing visualizations, as well as
plotting and graphing using Python libraries like Matplotlib, ggplot, bokeh, and
Pygal.

Table of Contents
Contents
Advanced Visualizations & Dashboards: Visualization Using Python..................................1
Table of Contents................................................................................................................... 1
Course Overview.................................................................................................................... 1
Relevance of Data Visualization for Business.....................................................................2
Libraries for Data Visualization in Python............................................................................3
Python Data Visualization Environment Configuration.......................................................4
Matplotlib Libraries for Visualization..................................................................................12
Bar Chart Using ggplot........................................................................................................13
Bokeh and Pygal.................................................................................................................. 17
Select Visualization Libraries..............................................................................................20
Interactive Graphs and Image Files....................................................................................22
Plot Graphs........................................................................................................................... 23
Multiple Lines in Graphs......................................................................................................25
Exercise: Create Line Charts with Pygal............................................................................28

Course Overview
[Video description begins] Topic title: Course Overview. [Video description
ends]

Hi, my name is Niranjan Pandey, and I'm part of Alchemy Software Solutions
e-learning and training content development team.

[Video description begins] Your host for this session is Niranjan Pandey. He is

a Mentor and a Solution Architect. [Video description ends]
I have got over 20 years of experience providing client-based training and e-
learning solutions. And, during my career, I have worked as developer, IBM
SOA architect, DevOps engineer, Hadoop administrator, and solution
architect. My IT certifications includes Java, IBM, Oracle, and Cloudera
Hadoop.

I specialize in development, system architecture, IT security, database design

and big data solution, virtualization and cloud solution, machine and deep
learning solution, automation and robotics, DevOps, and IoT. Data
visualization is an essential requirement for data exploration and it helps
businesses evaluate their progress and accordingly make appropriate
predictions and decisions.

Data visualization is a method of communicating information using visual

representations. It provides a way to tell a story about the data to a particular
audience. Visualizing information allows an audience to achieve some insight
or understanding about what the data suggests.

In this course, I will explore how to implement advanced visualization, plot and
graph using Python. I will discuss the scenario and approach of building and
implementing visualization plots and graphs using essential libraries of Python
that includes Matplotlib, ggplot, Bokeh, and Pygal.

Relevance of Data Visualization for Business

[Video description begins] Topic title: Relevance of Data Visualization for
Business. Your host for this session is Niranjan Pandey. [Video description
ends]

Objective of this video is to recognize the importance and relevance of data

visualization from business perspective.

[Video description begins] Screen title: Data Visualization Benefits. [Video

description ends]

Business process is driven by the data. If data is handled well, it will assist
business in order to grow by providing different calculations, manipulations of
data. Data visualization is one of the techniques in order to elaborate nature of
data and also provide insight in the business. And it does it by enabling
decision makers to take swift action after analyzing the data through
visualizations. It also provides capability to decision makers to identify key
business trends and base their decisions for future forecast on that.
Apart from that, business decision makers and business drivers would be able
to interpret the data quick, swift, and also concise. Data visualization
facilitates direct and customized interaction with data, providing summarized
data with good visual depictions, and also helps in doing forecasting and
predictions. Once data is understood by various different stakeholders, it
directly impacts productivity and also turns productivity into better sales. Now
let's try to understand relevance of data visualization for businesses.

It allows businesses to clearly understand the data, trends of data, identify

patterns, and also changes along with the driver of the changes. It allows
users to understand large amount of data through intuitive graphical
representations. Apart from that, it also helps in discovery of relationships and
pattern which exist in data and information. If you have used visualizations
which are intuitive, you can simplify the data interaction as well. Now let's try
to understand some practical benefit of data visualizations which can be
utilized by decision makers.

First one is by using data visualization and information that it depicts,

decisions can be made better. It provides capability of coming up with
enhanced analytical benefits which provides analytics to prepare ad hoc
reports, and also visualize trends in real-time data. It makes data available to
most of the users, and each user can go and customize the visualization as
per their use. This is what is called as self-serving reporting capabilities.

Apart from that, if you provide data in concise way, it increases productivity
and, above all, it saves lots of time of data manipulation. If you have data, in
form of visualization, gathering ideas from the team in order to identify the
gaps and create a shared view to align the decision-making is easier. Now
let's try to understand some of the visualization types.

[Video description begins] Screen title: Important Visualization Use

Case. [Video description ends]

You have line chart, where you can depict how actually growth has been there
in the identified parameter over time. Then you have gauge which tells you the
current relative business development. You can also have vertical or
horizontal bar graph and you can have pie graph which provides you
capability of identifying share of different components within a single chart.

Libraries for Data Visualization in Python

[Video description begins] Topic title: Libraries for Data Visualization in
Python. Your host for this session is Niranjan Pandey. [Video description
ends]

Objective of this video is to list various essential libraries that we can use in
Python to implement data visualization.

[Video description begins] Screen title: Libraries of Python [Video description

ends]

Python is one of the useful tools which is used by data scientists. It provides
you capability of data curating. It provides you capability of data extraction. It
provides you capability of data elaboration, and also provides you capability of
data wrangling, along with rich visualization. Some of the important libraries of
Python are Pandas. Pandas provides you capability of data manipulation and
analysis. Seaborn provides you statistical graphical representations. Pygal
provides your capability of creating SVGs.

Then we have Plotly, which helps you to plot different type of charts and
graphs, and it can be utilized by various languages. Then you have Bokeh,
which is an interacting visualization libraries and targets modern web
browsers for presentation. Let's understand Pandas in greater detail. Pandas
is highly optimized, built-in options for plotting visualizations. Apart from that,
name itself is derived from panel data, which provides you capability of
organizing the data in form of datasets, and do your observations over
multiple time periods for the same individual dataset. It integrates well with
other libraries of Python.

Then you have Seaborn, which is used for statistical graphics. It's an open
source software library, provides capability of building statistical graphics on
top of Matplotlib, and also closely integrates with Pandas. Apart from that, it
comes with built-in themes for styling the information for better visibility and
rich user interface for plotting the data. It also provides you capability of
drawing complicated plots in simple way. Then you have Bokeh, which is
meant for real-time interactive data presentation, which can be presented on
modern web browser.

It also provides customization options and functionalities. Then you have

Pygal. Pygal specializes in allowing the users to create SVGs. It is part of
Python library, open-source, and provides customization options along with
simplicity. You can go and you can embed Pygal charts and graphs and plots
into a web page. Apart from that, you have Plotly, which is an interactive
online visualization tool, comes with great API written in various different
languages. It can be easily modified clicking on different parts, provides you
capability of drill down and drill up charts and graphs.

Python Data Visualization Environment Configuration

[Video description begins] Topic title: Python Data Visualization Environment
Configuration. Your host for this session is: Niranjan Pandey. [Video
description ends]

Objective of this video is to set up data visualization environment using

Python tools and libraries. In order to ensure that we have collection of Python
tools and libraries installed, we can download and install Anaconda.

[Video description begins] An Anaconda Python/R Distribution page opens up

in a browser. [Video description ends]

It's open source, easiest possible platform that can be utilized to perform
Python and R data science and machine learning on various different
platforms. There are multiple different packages. Along with packages, it gives
a package manager tool called as Conda. It also facilitates in developing
machine learning and deep learning models using various different libraries.

Apart from that, we can also utilize visualizations using libraries like Matplotlib,
Bokeh, Datashader, HoloViews, and others. Let's go and download Anaconda
from www.anaconda.com/distribution and clicking on Download.

[Video description begins] He points to the url of the page. It reads:

https://www.anaconda.com/distribution/. He clicks the Download button
present at the top of the page. A new page opens. It has three tabs: Windows,
macOS, and Linux. Currently, the macOS tab is open. It has the heading:
Anaconda 2018.12 for macOS Installer. Two versions are displayed: Python
3.7 and Python 2.7. Each version has a Download button. [Video description
ends]

Once you'll click on Download, it will provide you distribution in different

platforms, and also two specific versions. We need to go and we need to
ensure that we have selected Windows.

[Video description begins] He clicks the Windows tab. The page heading is
Anaconda 2018.12 for Windows Installer. Two versions are displayed: Python
3.7 and Python 2.7. Each version has a Download button. [Video description
ends]

And once you get Anaconda 2018.12 for Windows Installer, we'll click on
Download.

[Video description begins] He clicks the Download button of Python 3.7

version. A dialog box appears. It displays the text: You have chosen to open:
Anaconda3-2018.12- Windows-x86_64.exe. There are two buttons at the
bottom: Save File and Cancel. [Video description ends]

Once you'll click on Download, it will let you download exe, click on Save File.
Wait for download to get completed. Once it is downloaded, you'll just go and
double-click on the download file.

[Video description begins] A File Explorer opens. The Downloads folder is

open. It has the downloaded file. A pop-up box displays with the following
message: Please wait while Setup is loading. [Video description ends]

It starts installation, and as step 1, it launches the Welcome screen.

[Video description begins] The File Explorer closes. An Anaconda Setup

window appears. It has two buttons: Next and Cancel. [Video description
ends]

You'll go and you'll click on Next of the wizard.

[Video description begins] The second step of the Setup is displayed, which

contains the License Agreement. Three buttons are displayed at the bottom:
Back, I Agree, and Cancel. [Video description ends]

In case, if it asks for license agreement, we'll say, I Agree.

[Video description begins] He clicks the I Agree button. The third step of the
Setup is displayed: Select Installation Type. Two options with radio buttons
are displayed for Install for: Just Me and All Users. At the bottom Back, Next,
and Cancel buttons are displayed. [Video description ends]

And finally, you can decide whether this installation will be supported to one
user or all users. We'll select All Users and click on Next. Once you'll click on
Yes, it will ask you to select the destination folder.
[Video description begins] The fourth step of the Setup: Choose Install
Location is displayed. It has an input box for file destination. A browse button
is placed next to it. At the bottom there are three buttons: Back, Next, and
Cancel. [Video description ends]

Let's go and select the destination folder. We'll go to Browse, and select
destination of our choice.

[Video description begins] He clicks the Browse button. A file explorer titled
Browse For Folder appears. It has a list of folders. At the bottom there are
three buttons: Make New Folder, OK, and Cancel. He selects the folder
named anaconda in the D Drive of the PC. He clicks the OK button. The file
explorer closes. The fourth step of the Setup window is open. The Destination
Folder is set as: D:\anaconda. [Video description ends]

We have created a folder called D:\anaconda, we'll click on Next.

[Video description begins] The fifth step of the Setup: Advanced Installation

Options appears. Check-boxes are displayed for two options: Add Anaconda
to the system PATH environment variable and Register Anaconda as the
system Python 3.7. The second option is selected by default. At the bottom,
there are three buttons: Back, Install, and Cancel. [Video description ends]

Then you can go and you can select whether you want to add Anaconda to
the path, we'll say Yes, though it is not recommended.

[Video description begins] He selects the first option. Both the options are
now selected. [Video description ends]

Finally, we'll click on Install.

[Video description begins] The Setup box heading reads: Installing. A

progress bar is displayed. A Show details button is present below it. At the
bottom, there are three buttons: Back, Next, and Cancel. [Video description
ends]

Once you'll click on Install, it will start the installation process, in that process,
it will extract various libraries and DLL files.

[Video description begins] He clicks the Show details button. It expands to

display a list of libraries and files. [Video description ends]
Once you get message indicating installation complete, it means all set of files
are downloaded and they are executed in order to help you in building
environment. We'll click on Next.

[Video description begins] The next heading reads: Microsoft Visual Studio

Code Installation. It has four buttons: Install Microsoft VSCode, Back, Skip,
and Cancel. [Video description ends]

You can also go and install Microsoft VSCode, if you wish to use it as your
editor, we'll say Skip for now.

[Video description begins] He clicks the Skip button. The final page of the
Setup is displayed. The heading is: Thanks for installing Anaconda3! At the
bottom, Back, Finish, and Cancel buttons are displayed. [Video description
ends]

And finally, we'll click on Finish.

[Video description begins] The Setup box closes. [Video description ends]

Once installation is successful, it will install various tools. One of the tool is
Spyder, which provides you capability of writing application, testing application
and also visualizing graphs if you are using appropriate libraries.

[Video description begins] A box titled spyder appears. [Video description

ends]

We'll wait for Spyder to get launched.

[Video description begins] A Windows Security Alert box appears. At the

bottom, there are two buttons: Allow access and Cancel. [Video description
ends]

You have to go and you have to allow access. And finally, you will get Spyder
which is divided into multiple different parts.

[Video description begins] He clicks the Allow access button. The Windows
Security Alert box closes. Spyder (Python 3.7) IDE opens on the screen. It
has a menu bar with options for File, Edit, Search, etc. It has a tool bar with
options for New, Open, Save, etc. The Spyder screen has three sections. The
first section is the Editor, in which the file untitled0.py is open. Next to it, there
is a second section which is the Console. It has a search bar and a Usage
box. Below it, there is a third section with the following three tabs: Help,
Variable explorer, and File explorer. The Help tab is open. It has a tab titled
Console 1/A . [Video description ends]

You have editor with one file called untitled0. On right hand side, you have
two panels, one is console. Another one displays variables under Variable
explorer and files under File explorer. You can click on Run to run the file and
see the outcome in console.

[Video description begins] Topic title: Python Data Visualization Environment

Configuration. Your host for this session is: Niranjan Pandey. [Video
description ends]

Objective of this video is to set up data visualization environment using

Python tools and libraries. In order to ensure that we have collection of Python
tools and libraries installed, we can download and install Anaconda.

[Video description begins] An Anaconda Python/R Distribution page opens up

in a browser. [Video description ends]

It's open source, easiest possible platform that can be utilized to perform
Python and R data science and machine learning on various different
platforms. There are multiple different packages. Along with packages, it gives
a package manager tool called as Conda.

It also facilitates in developing machine learning and deep learning models

using various different libraries. Apart from that, we can also utilize
visualizations using libraries like Matplotlib, Bokeh, Datashader, HoloViews,
and others. Let's go and download Anaconda from
www.anaconda.com/distribution and clicking on Download.

[Video description begins] He points to the url of the page. It reads:

https://www.anaconda.com/distribution/. He clicks the Download button
present at the top of the page. A new page opens. It has three tabs: Windows,
macOS, and Linux. Currently, the macOS tab is open. It has the heading:
Anaconda 2018.12 for macOS Installer. Two versions are displayed: Python
3.7 and Python 2.7. Each version has a Download button. [Video description
ends]

Once you'll click on Download, it will provide you distribution in different

platforms, and also two specific versions. We need to go and we need to
ensure that we have selected Windows.
[Video description begins] He clicks the Windows tab. The page heading is
Anaconda 2018.12 for Windows Installer. Two versions are displayed: Python
3.7 and Python 2.7. Each version has a Download button. [Video description
ends]

And once you get Anaconda 2018.12 for Windows Installer, we'll click on
Download.

[Video description begins] He clicks the Download button of Python 3.7

version. A dialog box appears. It displays the text: You have chosen to open:
Anaconda3-2018.12- Windows-x86_64.exe. There are two buttons at the
bottom: Save File and Cancel. [Video description ends]

Once you'll click on Download, it will let you download exe, click on Save File.
Wait for download to get completed. Once it is downloaded, you'll just go and
double-click on the download file.

[Video description begins] A File Explorer opens. The Downloads folder is

open. It has the downloaded file. A pop-up box displays with the following
message: Please wait while Setup is loading. [Video description ends]

It starts installation, and as step 1, it launches the Welcome screen.

[Video description begins] The File Explorer closes. An Anaconda Setup

window appears. It has two buttons: Next and Cancel. [Video description
ends]

You'll go and you'll click on Next of the wizard.

[Video description begins] The second step of the Setup is displayed, which

contains the License Agreement. Three buttons are displayed at the bottom:
Back, I Agree, and Cancel. [Video description ends]

In case, if it asks for license agreement, we'll say, I Agree.

[Video description begins] He clicks the I Agree button. The third step of the
Setup is displayed: Select Installation Type. Two options with radio buttons
are displayed for Install for: Just Me and All Users. At the bottom Back, Next,
and Cancel buttons are displayed. [Video description ends]
And finally, you can decide whether this installation will be supported to one
user or all users. We'll select All Users and click on Next. Once you'll click on
Yes, it will ask you to select the destination folder.

[Video description begins] The fourth step of the Setup: Choose Install

Location is displayed. It has an input box for file destination. A browse button
is placed next to it. At the bottom there are three buttons: Back, Next, and
Cancel. [Video description ends]

Let's go and select the destination folder. We'll go to Browse, and select
destination of our choice.

[Video description begins] He clicks the Browse button. A file explorer titled
Browse For Folder appears. It has a list of folders. At the bottom there are
three buttons: Make New Folder, OK, and Cancel. He selects the folder
named anaconda in the D Drive of the PC. He clicks the OK button. The file
explorer closes. The fourth step of the Setup window is open. The Destination
Folder is set as: D:\anaconda. [Video description ends]

We have created a folder called D:\anaconda, we'll click on Next.

[Video description begins] The fifth step of the Setup: Advanced Installation

Options appears. Check-boxes are displayed for two options: Add Anaconda
to the system PATH environment variable and Register Anaconda as the
system Python 3.7. The second option is selected by default. At the bottom,
there are three buttons: Back, Install, and Cancel. [Video description ends]

Then you can go and you can select whether you want to add Anaconda to
the path, we'll say Yes, though it is not recommended.

[Video description begins] He selects the first option. Both the options are
now selected. [Video description ends]

Finally, we'll click on Install.

[Video description begins] The Setup box heading reads: Installing. A

progress bar is displayed. A Show details button is present below it. At the
bottom, there are three buttons: Back, Next, and Cancel. [Video description
ends]

Once you'll click on Install, it will start the installation process, in that process,
it will extract various libraries and DLL files.
[Video description begins] He clicks the Show details button. It expands to
display a list of libraries and files. [Video description ends]

Once you get message indicating installation complete, it means all set of files
are downloaded and they are executed in order to help you in building
environment. We'll click on Next.

[Video description begins] The next heading reads: Microsoft Visual Studio

Code Installation. It has four buttons: Install Microsoft VSCode, Back, Skip,
and Cancel. [Video description ends]

You can also go and install Microsoft VSCode, if you wish to use it as your
editor, we'll say Skip for now.

[Video description begins] He clicks the Skip button. The final page of the
Setup is displayed. The heading is: Thanks for installing Anaconda3! At the
bottom, Back, Finish, and Cancel buttons are displayed. [Video description
ends]

And finally, we'll click on Finish.

[Video description begins] The Setup box closes. [Video description ends]

Once installation is successful, it will install various tools. One of the tool is
Spyder, which provides you capability of writing application, testing application
and also visualizing graphs if you are using appropriate libraries.

[Video description begins] A box titled spyder appears. [Video description

ends]

We'll wait for Spyder to get launched.

[Video description begins] A Windows Security Alert box appears. At the

bottom, there are two buttons: Allow access and Cancel. [Video description
ends]

You have to go and you have to allow access. And finally, you will get Spyder
which is divided into multiple different parts.

[Video description begins] He clicks the Allow access button. The Windows
Security Alert box closes. Spyder (Python 3.7) IDE opens on the screen. It
has a menu bar with options for File, Edit, Search, etc. It has a tool bar with
options for New, Open, Save, etc. The Spyder screen has three sections. The
first section is the Editor, in which the file untitled0.py is open. Next to it, there
is a second section which is the Console. It has a search bar and a Usage
box. Below it, there is a third section with the following three tabs: Help,
Variable explorer, and File explorer. The Help tab is open. It has a tab titled
Console 1/A . [Video description ends]

You have editor with one file called untitled0. On right hand side, you have
two panels, one is console. Another one displays variables under Variable
explorer and files under File explorer. You can click on Run to run the file and
see the outcome in console.

Matplotlib Libraries for Visualization

[Video description begins] Topic title: Matplotlib Libraries for Visualization.
Your host for this session is Niranjan Pandey. Screen title: Matplotlib. [Video
description ends]

Objective of this video is to list the prominent data visualization libraries that
we can use with Matplotlib. Matplotlib is a typical plotting library, which is used
by Python developers. It has numerical mathematical extensions which are
provided by Python. For example, it supports NumPy. It also provides an
object oriented API. It is considered as oldest and most popular InfoVis library.
Provides extensive range of 2D plot types and output formats.

It also provides support to 3D visualizations. Now let's try to understand some

of the libraries which are based on Matplotlib. First one is ggplot. ggplot and
ggplot2 is a data visualization package which is meant for statistical
programming languages. Second important library is plotnine. It's an
implementation of grammar of graphics in Python and it is based on ggplot.
This grammar allows user to compose plots by explicitly mapping data to the
visual object that makes up the plot. Then, you have HoloViews.

HoloViews is an open-source Python library designed to make data analysis

and visualizations simple. Finally, you have GeoViews. It's a Python library
that makes it easy to explore and visualize geographical and oceanographic
data sets, such as those which are used with weather, climate, remote
sensing research. There are two different type of APIs that comes up with
Matplotlib. Pyplot API and Object-oriented API. Pyplot API combines Pyplot
with Numpy into a single namespace, whereas, when we talk about object-
oriented API, Matplotlib itself is object-oriented.
And it is recommended to work with the objects directly if you want more
control and customizations of your plot. Now let's talk about some of the
toolkits which comes along with Matplotlib. We have mplot3d API. Basic
intention was to allow user to create simple 3D graphics with similar look and
feel as Matplotlib2d plots. Then you have Matplotlib axes_grid1 Toolkit. It was
actually collection of helper classes to simplify displaying multiple images in
Matplotlib.

Then you have Matplotlib axisartist Toolkit. It included a derived axis

implementation. And the biggest difference is that artists that are responsible
for drawing axis lines, ticks, tickslable, and axis labels are separated out from
MPI axis class. Finally, you have Matplotlib access toolkit. It's again a
collection of helper classes to simplify displaying multiple images in Matplotlib.

Bar Chart Using ggplot

[Video description begins] Topic title: Bar Chart Using ggplot. Your host for
this session is Niranjan Pandey. [Video description ends]

Objective of this video is to demonstrate the steps involved in creating bar

chart using ggplot in Python.

[Video description begins] A Colaboratory web page titled Untitled9.ipynb is

open in a browser. At the top, a menu bar is present with the following
options: File, Edit, View, etc. Below it, four buttons are shown: + Code, + Text,
Upward Cell, and Downward Cell. Below the buttons, a code editor is
displayed. Two code cells are shown in the code editor. The first code cell has
a code line. It reads: !pip install ggplot. The second code cell has four lines of
code. A play icon is present in each code cell. [Video description ends]

In order to ensure that we're able to utilize capabilities of ggplot typically to

create a bar chart, we need to first understand what is source of our data. We
are using Google Colab in order to prepare the environment. In the first cell,
we have written pip install ggplot. After writing this, we will go and we'll
execute it.

[Video description begins] He hits the play icon to execute the code. [Video
description ends]

Once you'll start execution, it will ensure that it installs ggplot and makes it
available for your other cells within the same session.
[Video description begins] It starts installing the required packages in the code
cell. [Video description ends]

Our next objective is now to go and prepare a bar chart using ggplot.

[Video description begins] In the second code cell, the first line reads: from
ggplot import *, the second line reads: import pandas as pd, the third line
reads: df = pd.DataFrame({"data1": [1,2,3,4], "data2": [10,30,40,50] }), and the
last line reads: ggplot(aes(x="data1", weight="data2"), df) +
geom_bar_(fill="#FF6666"). [Video description ends]

We are first importing capabilities of ggplot by writing ggplot import *.

[Video description begins] He points to the first code line in the second code
cell. [Video description ends]

Then we are importing pandas as pd in order to manipulate data.

[Video description begins] He points to the second code line in the second
code cell. [Video description ends]

We are taking here static data in order to build Pandas DataFrame, we have
written df = pd.DataFrame, and we are creating data1 and data2 as two
different datasets.

[Video description begins] He points to the third code line in the second code
cell. [Video description ends]

These two datasets will help in creating dataframe. Once dataframe is

created, we'll use ggplot, pass the axis, specify X axis to be one dataset that
is data1, weight to be another dataset, that is data2. And we'll pass the
dataframe that is df, which is initialized in the previous line.

[Video description begins] He points to the fourth code line in the second code
cell. [Video description ends]

Then we'll go and we'll specify some of the properties, for example, we want
this graph to be filled by a particular color. You can use the color in order to fill
the graph. After you have written, you have to go and you have to execute the
current cell.
[Video description begins] He hits the play icon to execute the code. A graph
with data2 on y-axis and data1 on x-axis appears in the code cell. [Video
description ends]

Once you'll execute the current cell, you'll find that your data2 is present in Y
axis. It starts with 10, 20, 30, 40, 50. And in X axis, you have data1, which is
1, 2, 3, 4. And you can clearly see that whatever color you have applied in fill
inside geom_bar, it is displaying your bar chart in that particular color.

[Video description begins] Topic title: Bar Chart Using ggplot. Your host for
this session is Niranjan Pandey. [Video description ends]

Objective of this video is to demonstrate the steps involved in creating bar

chart using ggplot in Python.

[Video description begins] A Colaboratory web page titled Untitled9.ipynb is

open in a browser. At the top, a menu bar is present with the following
options: File, Edit, View, etc. Below it, four buttons are shown: + Code, + Text,
Upward Cell, and Downward Cell. Below the buttons, a code editor is
displayed. Two code cells are shown in the code editor. The first code cell has
a code line. It reads: !pip install ggplot. The second code cell has four lines of
code. A play icon is present in each code cell. [Video description ends]

In order to ensure that we're able to utilize capabilities of ggplot typically to

create a bar chart, we need to first understand what is source of our data. We
are using Google Colab in order to prepare the environment. In the first cell,
we have written pip install ggplot. After writing this, we will go and we'll
execute it.

[Video description begins] He hits the play icon to execute the code. [Video
description ends]

Once you'll start execution, it will ensure that it installs ggplot and makes it
available for your other cells within the same session.

[Video description begins] It starts installing the required packages in the code
cell. [Video description ends]

Our next objective is now to go and prepare a bar chart using ggplot.

[Video description begins] In the second code cell, the first line reads: from
ggplot import *, the second line reads: import pandas as pd, the third line
reads: df = pd.DataFrame({"data1": [1,2,3,4], "data2": [10,30,40,50] }), and the
last line reads: ggplot(aes(x="data1", weight="data2"), df) +
geom_bar_(fill="#FF6666"). [Video description ends]

We are first importing capabilities of ggplot by writing ggplot import *.

[Video description begins] He points to the first code line in the second code
cell. [Video description ends]

Then we are importing pandas as pd in order to manipulate data.

[Video description begins] He points to the second code line in the second
code cell. [Video description ends]

We are taking here static data in order to build Pandas DataFrame, we have
written df = pd.DataFrame, and we are creating data1 and data2 as two
different datasets.

[Video description begins] He points to the third code line in the second code
cell. [Video description ends]

These two datasets will help in creating dataframe. Once dataframe is

created, we'll use ggplot, pass the axis, specify X axis to be one dataset that
is data1, weight to be another dataset, that is data2. And we'll pass the
dataframe that is df, which is initialized in the previous line.

[Video description begins] He points to the fourth code line in the second code
cell. [Video description ends]

Then we'll go and we'll specify some of the properties, for example, we want
this graph to be filled by a particular color. You can use the color in order to fill
the graph. After you have written, you have to go and you have to execute the
current cell.

[Video description begins] He hits the play icon to execute the code. A graph
with data2 on y-axis and data1 on x-axis appears in the code cell. [Video
description ends]

Once you'll execute the current cell, you'll find that your data2 is present in Y
axis. It starts with 10, 20, 30, 40, 50. And in X axis, you have data1, which is
1, 2, 3, 4. And you can clearly see that whatever color you have applied in fill
inside geom_bar, it is displaying your bar chart in that particular color.
Bokeh and Pygal
[Video description begins] Topic title: Bokeh and Pygal. Your host for this
session is Niranjan Pandey. [Video description ends]

Objective of this video is to demonstrate how to create charts using libraries of

Bokeh and Pygal in Python. In order to implement Python, we are using Colab
which comes with Google.

[Video description begins] A Colaboratory editor titled 1_07.ipynb is open in a

browser. It has a menu bar with the following options: File, Edit, View, etc.
Below it, four buttons are shown: + Code, +Text, Upward Cell, and Downward
Cell. Below the buttons, a code editor is displayed. It has a Run icon at left
and a few code cells. [Video description ends]

In order to ensure that we are able to do research without hindrance created

by hardware requirement and software requirement. First, we are taking up
Pygal, which provides us capability of generating charts and graphs which can
be file-based.

We can store it in a particular file or we can generate SVG images. In order to

utilize Pygal, we are importing display, import HTML in order to ensure that
content can be displayed in HTML format. Then we are importing Pygal.

[Video description begins] The presenter is currently in the first code cell. He

talks about code lines 1 and 2. Code line 1 reads: from IPython.display import
HTML. Code line 2 reads: import pygal. [Video description ends]

We have created directive where we have specified doctype HTML and HTML
opening and closing.

[Video description begins] He talks about code lines 5 and 6. Code line 5 is: <!
DOCTYPE html>. Code line 6 is: <html>. [Video description ends]

Important element here is body figure, where we are specifying a regular

expression called as pygal_render, which we are creating in below lines.

[Video description begins] He talks about code line 11. It reads

<body><figure>{pygal_render}</figure></body>. [Video description ends]

Chart construction start when we initialize line_chart = pygal.Line.

[Video description begins] He talks about code line 15. It reads: line_chart =
pygal.Line(). [Video description ends]

Then we are specifying what should be the title of the chart we are using,
Product usage evolution in percentage. Then we are providing labels which is
map ( str, which ranges from 2002 to 2013), which is the year when we are
going to display the product evolution.

[Video description begins] He talks about code line 16. It reads: line_chart.title
= "Product usage evolution (in %)". [Video description ends]

[Video description begins] He talks about code line 17. It reads:

line_chart.x_labels = map(str, range(2002, 2013)). [Video description ends]

Products that we have added to our chart are product 1, product 2, product 3,
and miscellaneous along with the data.

[Video description begins] He talks about code lines 18 to 21. Code line 18 is:
line_chart.add("Product1", [None, None, 0, 16.6, 25, 31, 36.4, 45.5, 46.3,
42.8, 37.1]). Code line 19 is: line_chart.add("Product2", [None, None, None,
None, None, None, 0, 3.9, 10.8, 23.8, 35.3]). Code line 20 is:
line_chart.add("Product3", [85.8, 84.6, 84.7, 74.5, 66, 58.6, 54.7, 44.8, 36.2,
26.6, 20.1]). Code line 21 is: line_chart.add("MISC", [14.2, 15.4, 15.3, 8.9, 9,
10.4, 8.9, 5.8, 6.7, 6.8, 7.5]). [Video description ends]

Finally, we are using HTML in order to ensure that we are able to launch the
chart specifying pygal_render, which will be utilized in the HTML directive that
we have specified in the above lines. And then finally, we are making call to
.render passing is_unicode=True.

[Video description begins] He talks about code line 22. It reads:

HTML(html_pygal . format(pygal_render=line_chart . render(is_unicode=True)
) ). [Video description ends]

When we run this cell, it will ensure that it generates a line chart utilizing
capabilities of Pygal.

[Video description begins] He clicks the Run icon. [Video description ends]

Apart from this, you can also go and store this chart in SVG. For that, yes, you
have to have a different function of render which is render_2_5. So depending
on your need, you can go and you can manipulate, evaluate and store Pygal
charts.

[Video description begins] A graph appears in the output. It is titled: Product

usage evolution (in %). The x-axis ranges from 2002 to 2012 and the y-axis
ranges from 10 to 80. In the graph, Product1 is marked in red, Product2 is
marked in purple, Product3 is marked in green, and MISC is marked in
yellow. [Video description ends]

Now let's go and try to evaluate Bokeh.

[Video description begins] He scrolls down the page and goes to the third
code cell. [Video description ends]

Bokeh provides capability of rendering, again, in form of HTML. If you want to

build web applications and you want applications to have charts which are
interactive, you can go and you can use Bokeh. In cell number 3, we have
imported numpy as np because we are using NumPy in order to generate the
data which we will be plotting.

[Video description begins] He talks about code line 1. It reads: import numpy
as np. [Video description ends]

Then we are taking figure and show, which is present in bokeh.plotting. And
finally, we are importing output_notebook because we want the chart to be
displayed in the notebook itself.

[Video description begins] He talks about code lines 2 and 3. Code line 2 is:
from bokeh.plotting import figure, show. Code line 3 is: from bokeh.io import
output_notebook. [Video description ends]

We are specifying n = 100 and then we are generating sequence of random

numbers and we are storing it in x and y.

[Video description begins] He talks about code lines 4 to 6. Code line 4 is: N =
100. Code line 5 is: x = np.random.random(size=N) * 50. Code line 6 is: y =
np.random.random(size=N) * 50. [Video description ends]

Finally, we are calculating radii, which is calculated using np.random.random

passing size equal to N multiplying it by 1.5.
[Video description begins] He talks about code line 7. It reads: radii =
np.random.random(size=N) * 1.5. [Video description ends]

Then we are going and selecting the colors that we want in order to plot the
chart.

[Video description begins] He talks about code line 8. It reads: colors = ["#
%02x%02x%02x" % (r, g, 150) for r, g in zip (np.floor(50+2*x) .astype (int),
np.floor(30+2*y) .astype(int))]. [Video description ends]

And that chart is getting plotted by using p = figure, p.circle specifying x and y
which are the data points. Radius equal to radii, which we have calculated
using np.random function. Fill color, we are using colors, which we are
passing as an array in colors variable. And finally, fill alpha equal to 0.6 and
line color equal to none.

[Video description begins] He talks about code lines 10 to 12. Code line 10
reads: p = figure(). Code line 11 reads: p.circle(x, y, radius=radii,
fill_color=colors, fill_alpha=0.6, line_color=None). Code line 12 reads:
show(p). [Video description ends]

Show does the magic of embedding your current chart which will be created
using Bokeh to the notebook. In order to display the chart in notebook, we'll go
and we'll execute the cell.

[Video description begins] He clicks the Run icon. [Video description ends]

Once you'll execute the cell, you'll find that it loads Bokeh.js and provides the
browser view of your chart.

[Video description begins] A graph appears in the output below the code lines.
It has the following text: BokehJS 1.0.4 successfully loaded. There is a list of
icons present on the right side of the graph. These icons are Save, Reset,
Refresh etc. [Video description ends]

You can go and you can save it. You can also refresh, so that in case if there
is change in data, that change is reflected in your chart as well.

Select Visualization Libraries

[Video description begins] Topic title: Select Visualization Libraries. Your host
for this session is Niranjan Pandey. Screen title: Visualization Selection Flow
Chart. [Video description ends]

Objective of this video is to recognize the essential selection criteria that we

have to consider when selecting appropriate data visualization library. One of
the tasks while deciding what all visualization capabilities you want on the
project, actually drives selection of tools. We are depicting a visualization
selection flow chart, where first you need to identify what's the nature of the
project? What is the domain?

Once you identify the domain, you need to go and find out whether UI/UX
design and customization is essential. If it is yes, better is to go and build
vector files which are SVG. But in case if UI/UX customization doesn't matter,
you need to go and identify whether resolution scale matters. If it matters, you
have to go and you have to select Canvas, which is Bitmap.

Then you need to focus on the volume of data and whether you have to
manage data at rest or real time data. Depending on the need, if you have to
manage big data or real time data, you can go and you can select Canvas
which is again, Bitmap. If you are selecting Bitmap or Canvas, you can use
ChartJS or you can use E-charts. If you are going with SVG, you can select
from Highcharts or Plotly.

[Video description begins] A flowchart appears. There is a Project that

connects to UI/UX customization via an arrow labelled Start. The UI/UX
customization is then connected to Vector file via an arrow labelled True and
to Resolution via an arrow labelled False. The Resolution further connects to
Big data or real-time data? via an arrow labelled True and to Canvas via an
arrow labelled False. The Big data or real-time data? further connects to
Canvas via an arrow labelled True and to Vector file via an arrow labelled
False. Screen title: Selecting Right Graphics Types. [Video description ends]

One of the task is to go and select the right graphics type, for example,
whether you should go for Bitmap or you should go for Vector. If your need is
high performance and you're working with filters, rays, and tracers, probably
you can go with Canvas.

But if you are working with high-fidelity documents, which are important for
viewing and printing, or you want to go with static images, probably SVG is
the better option.
[Video description begins] An illustration is shown. In it, two images
representing canvas and SVG are connected to each other via a double-sided
arrow. The arrow is labelled on one side as High performance and Image and
on the other side as Hi-fi format. Screen title: Chart Library Comparison
Matrix. [Video description ends]

Now let's try to understand various libraries and their capabilities through
comparison matrix. Features that we have selected here are graphics type,
ChartJS's Bitmap, Highcharts Vector, plotly Vector, Britecharts Vector. When
it comes to dependencies, ChartJS depends on color moment. Highchart have
no dependency.

Plotly.js depends on 50 modules, and Britecharts depends on D3 v4+. From

perspective of size of these libraries, you'll find that ChartJS is 154 KB.
Highcharts is 187, plotly.js is 2.23 MB and Britecharts is 1.96 MB. From
perspective of documentation available for APIs, ChartJS is bit confusing,
Highchart is good, Plotly.js is not as good as Britecharts and Highcharts.

From interface perspective, ChartJS and Highcharts are simple. Plotly.js is

difficult, you need some programming benefit. Britechart provides you
immediate interface. Licensing perspective, ChartJS is MIT licensed,
Highcharts comes with commercial and free, plotly.js is MIT and Brightcharts
are Apache 2.0 license compliant.

Interactive Graphs and Image Files

[Video description begins] Topic title: Interactive Graphs and Image Files.
Your host for this session is Niranjan Pandey. Screen title: Interactive Graph
Features. [Video description ends]

Objective of this video is to recall the scenarios of creating interactive graph

as against image files. Relevance of data depends on the timeline. If
relevance is based on time, probably interactive graph is better. But if you feel
that historical data also is relevant, you can go and you can have charts and
graphs which are static in nature and can also be stored in a file.

Let's go ahead and understand interactive graph. Interactive graph is a chart

on which when you hover a point, zoom, play with axis, add a button, and so
on. It has different advantages which are relevant from user perspective.
Apart from that it also enables displaying and understanding multi-dimensional
data, provides variety of visualization chart and enables traditional data
exploration task using interactive chart.
[Video description begins] Screen title: Interactive Graph Types. [Video
description ends]

Some of the interactive charts which are essential are heat map. Heat map is
a graphical representation of data where individual values contained in a
matrix are represented with different colors. Heat map is a newer term but
shading matrices has existed for long. Then we have geographical map.
Information which are related to geography can be represented using
geographical map.

For example, if you want to point out sales across different geographies
probably you will go with geographical map. Then you have link chart. Link
charts responsibility is to depict or show the magnitude and direction of
relationships between two or more categorical variables.

They are used in link analysis for identifying relationships between nodes that
are not easy to see from the existing raw data. Now let's try to understand
Image File Graphs. They are typical infographics which are posted in files.
Focuses on a specific data story rather than changing and providing multidata
stories in a single visualization. It provides single view for exploration.

Plot Graphs
[Video description begins] Topic title: Plot Graphs. Your host for this session
is Niranjan Pandey. [Video description ends]

Objective of this video is to demonstrate how to plot graphs using lines and
markers.

[Video description begins] An RStudio IDE opens. At the top, it has a menu
bar with the following options: File, Edit, Code, View, etc. The IDE has four
sections. The first section is the Code Editor. Inside the Code Editor, there is
a panel with a few buttons. Some of these are: Save, Source on Save, Run,
etc. A tab titled Untitled is open here. It has fourteen lines of code. The
second section is below the Code Editor. It has two tabs: Console and
Terminal. The Console tab is open. The third section lies on the right of the
Code Editor. It has three tabs: Environment, History, and Connections. The
Environment tab is open. Below this section lies the fourth section. It has five
tabs: Files, Plots, Packages, Help, and Viewers. The Plots tab is open. [Video
description ends]
We're using ggplot2 and reshape2 libraries which are part of R Studio or
which comes along with R. You can go and you can install package if it is not
already installed. After installing package, you have to go and get the library
available in your program. In line number 1 and 2, we are using the libraries
which we are going to use for this current program.

[Video description begins] Code line 1 reads: library(ggplot2). Code line 2

reads: library(reshape2). [Video description ends]

Then we are setting the seed value to 1.

[Video description begins] He talks about code line 4. It reads: set.seed <-
1. [Video description ends]

We are creating a data frame using data.frame. We are passing cbind. Cbind
typically is a function which is used in R that appends or combines vectors,
matrices or data frames by columns. We are passing sequence 1,10,1 and
then matrix along with rnorm. Rnorm is used in order to generate random
numbers with arguments mean and at standard deviation. It can be used in
order to generate uniform number.

[Video description begins] He points to code line 5. It reads: df <-

data.frame(cbind(seq(1, 10, 1),matrix(rnorm(100, 1, 20), 10, 4))). [Video
description ends]

Then, we are using melt df, we are passing id=x1. After passing id=x1
and initializing d, we will have sufficient data in order to go and plot the line.

[Video description begins] He points to code line 6. It reads: d <- melt(df,

id="X1"). [Video description ends]

In order to plot line with different styles, we are using ggplot. We are passing
the data. And we are specifying x-axis and y-axis along with color variable. In
other words, it will choose color on its own depending on the number of lines.

[Video description begins] He points to code line 9. It reads: ggplot(d,

aes(x=X1, y=value, color=variable))+. [Video description ends]

Then we are specifying line type is variable and size is 1. We'll execute this
code in order to see how actually lines will get created.
[Video description begins] He points to code line 10. It reads:
geom_line(aes(linetype=variable), size=1). [Video description ends]

After selecting the part of code we'll click on Run.

[Video description begins] He selects code lines 1 to 10 and hits the Run
button. The selected code lines appear in the console pane. [Video
description ends]

Once we'll click on Run, it will execute and it will plot the graph where it will
use variables as x2, x3, x4 and x5 along with different lines.

[Video description begins] A graph appears in the Plots tab. It has value on
the y-axis and X1 on the x-axis. Four variable lines are present in the graph.
These are: X2, X3, X4, and X5. [Video description ends]

Now our objective is to go and see what happens when we want to add
different markers for each line.

[Video description begins] He talks about code line 12. It reads: # Line Styles
with markers 2: different markers for each line. [Video description ends]

To do that we have written ggplot d. We are passing x = x1, y = value, color =

variable. And now we are adding geom_point aes, shape = variable, size = 4,
in line number 13 and 14. This will add the marker which will be added as a
result of line number 14.

[Video description begins] Code line 13 reads: ggplot(d, aes(x=X1, y=value,

color=variable))+. Code line 14 reads: geom_line() +
geom_point(aes(shape=variable, size=4))+. [Video description ends]

Let's go and execute the complete program now. We'll Select All and we'll
click on Run.

[Video description begins] He selects code lines 1 to 14 and clicks the Run
button. [Video description ends]

Once we'll click on Run, you'll find that it will generate two charts, and it will
bring up the most recent one up in your plotter.

[Video description begins] A graph appears in the Plots tab. It has value on
the y-axis and X1 on the x-axis. Four variable lines are present in the graph.
These are: X2, X3, X4, and X5. The pink line on the graph, labelled X2,
shows pink circles as markers at each new point in the line. The green line on
the graph, labelled X3, shows green triangles as markers at each new point in
the line. The blue line on the graph, labelled X4, shows blue rectangles as
markers at each new point in the line. The purple line on the graph, labelled
X5, shows purple cross signs as markers at each new point in the line. [Video
description ends]

You can see there are different shapes for different markers, triangle,
rectangle, and circle. But when you go to previous one, it shows you simple
line without any marker. This is as a result of line number 8 and 9.

[Video description begins] He presses the backward arrow button in the Plots
tab and shifts to the previous graph. [Video description ends]

And, second chart which comes with marker as a result of line number 13 and
14.

[Video description begins] He presses the forward arrow button in the Plots
tab and shifts to the new graph. [Video description ends]

Multiple Lines in Graphs

[Video description begins] Topic title: Multiple Lines in Graphs. Your host for
this session is Niranjan Pandey. [Video description ends]

Objective of this video is to plot several lines in a single graph using different
line styles and markers. In order to ensure that we are able to plot several
lines in a single graph, first task is to go and ensure that we have taken up the
libraries.

[Video description begins] An RStudio IDE opens. At the top, it has a menu
bar with the following options: File, Edit, Code, View, etc. The IDE has four
sections. The 1st section is the Code Editor. Inside the Code Editor, there is a
panel with a few buttons. Some of these are: Save, Source on Save, Run, etc.
A tab titled 1_11.R is open here. It has eleven lines of code. The 2nd section
is below the Code Editor. It has two tabs: Console and Terminal. The Console
tab is open. The 3rd section lies on the right of the Code Editor. It has three
tabs: Environment, History, and Connections. The Environment tab is open.
Below this section lies the 4th section. It has 5 tabs: Files, Plots, Packages,
Help, and Viewers. The Plots tab is open. [Video description ends]
In line number 1 and 2, we are taking library ggplot2 and reshape2.

[Video description begins] Code line 1 reads: library(ggplot2). Code line 2

reads: library(reshape2). [Video description ends]

In line number 4, we are setting the seed value. In line number 5, we are
generating random sequence and matrices, which we are binding in order to
build a data frame.

[Video description begins] Code line 4 reads: set.seed <- 1. Code line 5

reads: df <- data.frame(cbind(seq(1, 10, 1),matrix(rnorm(100, 1, 20), 10,
4))). [Video description ends]

After building data frame, we are deriving d dataset, which we will be using in
line number 9 in order to start plotting.

[Video description begins] Code line 9 reads: ggplot(d, aes(x=X1, y=value,

color=variable))+ . [Video description ends]

In line number 9, we are specifying ggplot function, we are passing d, we are

specifying axis, x = X1, y = value, and we are specifying color to be variable.
Plus now we are specifying geom_line, specifying linetype as variable and
size=1.

[Video description begins] He talks about code line 10. It reads:

geom_line(aes(linetype=variable), size=1). [Video description ends]

In line number 11, we are specifying marker, which will be marking the points
with different shape. We have taken size as 10 which controls the width of the
marker.

[Video description begins] Code line 11 reads:

geom_point(aes(shape=variable, size=10)). [Video description ends]

Let's go and execute it to derive the outcome. We'll select the entire code
which we have written in the script, and we'll click on Run button.

[Video description begins] He selects code lines 1 to 11 and clicks the Run
button. The selected code lines appear in the console pane. [Video
description ends]

Once we'll click on Run button, it will start plotting.

[Video description begins] A graph appears in the Plots tab. It has value on
the y-axis and X1 on the x-axis. Four variable lines are present in the graph
with size 1 and marker size 10. These are: X2, X3, X4, and X5. The pink line
on the graph, labelled X2, shows pink circles as markers at each new point in
the line. The green line on the graph, labelled X3, shows green triangles as
markers at each new point in the line. The blue line on the graph, labelled X4,
shows blue rectangles as markers at each new point in the line. The purple
line on the graph, labelled X5, shows purple cross signs as markers at each
new point in the line. [Video description ends]

You'll see that it generates different lines in a single graph along with different
style and marker. You'll find different colors. In chase if you want to increase
the size of the line, you can go and you can change the value from size 1 to
size 10. Re-select it, run the same script again.

[Video description begins] In code line 10, he changes the size to 10. Code
line 10 now reads: geom_line(aes(linetype=variable), size=10). He again
selects code lines 1 to 11 and clicks the Run button. [Video description ends]

Once you'll run the same script again, you'll find that line thickness changes. It
is only because we have changed the size of the line to 10.

[Video description begins] A graph appears in the Plots tab. It has value on
the y-axis and X1 on the x-axis. Four variable lines are present in the graph
with size 10 and marker size 10. These are: X2, X3, X4, and X5. The
thickness of the lines in the new graph has increased. The pink line on the
graph, labelled X2, shows pink circles as markers at each new point in the
line. The green line on the graph, labelled X3, shows green triangles as
markers at each new point in the line. The blue line on the graph, labelled X4,
shows blue rectangles as markers at each new point in the line. The purple
line on the graph, labelled X5, shows purple cross signs as markers at each
new point in the line. [Video description ends]

We'll revert back to 1, and we will change the size of marker now to 1.

[Video description begins] He reverts the changes in code line 10. In code line
11, he changes the size to 1. Code line 11 now reads:
geom_point(aes(shape=variable, size=1)). [Video description ends]

We'll select once again and we'll click on Run.

[Video description begins] He again selects code lines 1 to 11 and clicks the
Run button. [Video description ends]

Once we'll click on Run, you'll find that now linetype is 1 and you'll find that
marker has reduced its size.

[Video description begins] A graph appears in the Plots tab. It has value on
the y-axis and X1 on the x-axis. Four variable lines are present in the graph
with size 1 and marker size 1. These are: X2, X3, X4, and X5. The thickness
of the lines has decreased. The pink line on the graph, labelled X2, shows
pink circles as markers at each new point in the line. The green line on the
graph, labelled X3, shows green triangles as markers at each new point in the
line. The blue line on the graph, labelled X4, shows blue rectangles as
markers at each new point in the line. The purple line on the graph, labelled
X5, shows purple cross signs as markers at each new point in the line. [Video
description ends]

So depending on availability of data, which is continuous variable data, in our

case, as of now, we have to go and we have to select appropriate linetype
and shape to manage multiple styles and markers in the same plot.

Exercise: Create Line Charts with Pygal

[Video description begins] Topic title: Exercise: Create Line Charts with Pygal.
Your host for this session is Niranjan Pandey. [Video description ends]

Now it's time to test your knowledge and understanding of what we have
learnt in this course. So in this exercise, you will create line chart using Pygal
in Python. Create HTML directive to render the chart, and finally render Pygal
chart created using HTML directive. Now I want you to pause this video and
attempt all of the exercises and then come back to view the solution so you
understand how well you have attempt.

[Video description begins] A Solution banner appears. A Colaboratory editor

is open in a browser window. It is titled: Exerciseq1_dd.ipynb. At the top, a
menu bar is present with the following options: File, Edit, View, etc. Below it,
four buttons are shown: + Code, + Text, Upward Cell, and Downward Cell.
Below the buttons, there is a Code Editor. It has one code cell. The code cell
has a play button. [Video description ends]
In order to create Pygal chart, we need to ensure that we have declared the
required packages. We are declaring here IPython.display import HTML. So
that we are able to specify HTML directive, then we are importing pygal.

[Video description begins] He points to code line 1. It reads: from

IPython.display import HTML. He points to code line 2. It reads: import
pygal. [Video description ends]

And finally, we are taking up HTML directive, which starts with DOCTYPE
html opening and closing, and we are storing it in html_pygal.

[Video description begins] He points to code lines 4 to 12. Code line 4 is:
html_pygal = u""" . Code line 5 is: <!DOCTYPE html>. Code line 6 is: <html>.
Code line 7 is: <head>. Code line 8 is: <script type="text/javascript"
src="http://kozea.github.com/pygal.js/javascripts/svg.jquery.js"></script>.
Code line 9 is: <script type="text/javascript"
src="http://kozea.github.com/pygal.js/javascripts/pygal-tooltips.js"></script>.
Code line 10 is: < /head >. Code line 11 is:
<body><figure>{pygal_render}</figure></body>. Code line 12 is:
</html>. [Video description ends]

Our next objective is to go and build the line chart that we are doing by
specifying line_chart = pygal.Line(). We are specifying title to our chart and we
are adding label and various different elements which we want to plot.

[Video description begins] He points to code lines 15 to 21. Code line 15 is:
line_chart = pygal.Line(). Code line 16 is: line_chart.title = "Product usage
evolution (in %)". Code line 17 is: line_chart.x_labels = map(str, range(2002,
2013)). Code line 18 is: line_chart.add("Product1", [None, None, 0, 16.6, 25,
31, 36.4, 45.5, 46.3, 42.8, 37.1]). Code line 19 is: line_chart.add("Product2",
[None, None, None, None, None, None, 0, 3.9, 10.8, 23.8, 35.3]). Code line
20 is: line_chart.add("Product3", [85.8, 84.6, 84.7, 74.5, 66, 58.6, 54.7, 44.8,
36.2, 26.6, 20.1]). Code line 21 is: line_chart.add("MISC", [14.2, 15.4, 15.3,
8.9, 9, 10.4, 8.9, 5.8, 6.7, 6.8, 7.5]). [Video description ends]

After creating line_chart, our objective is to go and plot it in the HTML which
we have created earlier. For that, we are using HTML, we are passing the
name of the variable where we have stored the directive, which is html_pygal.
We are formatting it by applying .format and passing
pygal_render=line_chat.render, specifying is_unicode=True.
[Video description begins] He points to code line 22. It reads:
HTML(html_pygal . format(pygal_render=line_chart . render(is_unicode=True)
) ). [Video description ends]

Once you have done that, it means that you have already created the chart.
And you have specified statements which are required in order to render the
chart in the HTML segment of your current Google Colab, which supports
Python code using various cells. We'll go and we'll click on Run cell.

[Video description begins] He clicks the play button. [Video description ends]

Once you'll click on Run cell, it will start executing the statement. And finally,
you'll find that it renders the chart which is line_chart, specifying products and
miscellaneous. And also specifies the header on the top of it as product usage
evolution, along with various different charts that indicates different products.

[Video description begins] A graph output appears below the code cell. It is
titled: Product usage evolution (in %). The x-axis ranges from 2002 to 2012
and the y-axis ranges from 10 to 80. In the graph, Product1 is marked in red,
Product2 is marked in purple, Product3 is marked in green, and MISC is
marked in yellow. [Video description ends]