Final Report of Traineeship Program 2020

On

“VISUALIZING COVID19”

MEDTOUREASY

27th June 2020

 ACKNOWLDEGMENTS
The traineeship opportunity that I had with MedTourEasy was a great change for
learning and understanding the intricacies of the subject of Data Visualizations in Data
Analytics; and also, for personal as well as professional development. I am very obliged
for having a chance to interact with so many professionals who guided me throughout
the traineeship project and made it a great learning curve for me.

Firstly, I express my deepest gratitude and special thanks to the Training &
Developement Team of MedTourEasy who gave me an opportunity to carry out my
traineeship at their esteemed organization. Also, I express my thanks to the team for
making me understand the details of the Data Analytics profile and training me in the
same so that I can carry out the project properly and with maximum client satisfaction
and also for spearing his valuable time in spite of his busy schedule.

I would also like to thank the team of MedTourEasy and my colleagues who made the
working environment productive and very conducive.
 TABLE OF CONTENTS

Acknowledgments .................................... i

Abstract ..................................................... iii

Sr. No. Topic Page No.

1 Introduction
1.1 About the Company 6
1.2 About the Project 7
1.3 Objectives and Deliverables 9
2 Methodology
2.1 Flow of the Project 11
2.2 Use Case Diagram 12
2.3 Language and Platform Used 13
3 Implementation
3.1 Gathering Requirements and Defining Problem Statement 16
3.2 Data Collection and Importing 16
3.3 Designing Databases 17
3.4 Data Cleaning 19
3.5 Data Filtering 20
3.6 Prototyping - Power BI 22
3.7 Development of Dashboards 24
4 Sample Screenshots and Observations
4.1 COVID-19 Worldwide Tracker 29
4.2 COVID-19 Testing Tracker
4.3 COVID-19 India Tracker
6 Conclusion
7 Future Scope
8 References
 ABSTRACT
In December 2019, COVID-19 was first identified in the Wuhan region of China. The
World Health Organization declared COVID-19 as a Public Health Emergency of
International Concern (PHEIC) on 30th January 2020. By March 11, it categorized COVID-
19 outbreak as a pandemic. Within a period of two to three months it spread to 210
countries across the world.

According to WHO reports, COVID-19 is a severe acute respiratory syndrome which is

transmitted through respiratory droplets and contact routes. It can be contracted by direct
contact with the infected person or indirect contact through touch of infected surfaces and
materials. Due to the infectious nature of this disease, it has spread to 210 countries and
territories around the world and infected (confirmed) cases have crossed the 7 million mark
worldwide.

The total (cumulative) number of confirmed infected people has crossed 2.5 lakh mark
as on 11th June 2020. To tackle this disease, governments all over the world have taken
major steps to combat it and to mitigate its effects. Most of the countries have
followed policies of social distancing, travel controls, staggered and complete
lockdown and intensive testing.

But these steps have brought the entire world to a standstill with many restrictions in all
countries. The inability for people to go out for economic activities and other revenue
generating actions is having a major financial impact on the world. Hence it is imperative
for organizations to study and track the data of this pandemic.

Therefore, this project aims at collecting and analyzing wide variety of large data sets,
create intuitive and interactive dashboards for representing COVID-19 in order to gain
meaningful insights.
1.1 About the Company
MedTourEasy, a global healthcare company, provides you the informational
resources needed to evaluate your global options. MedTourEasy provides
analytical solutions to our partner healthcare providers globally.

1.2 About the Project

The 2019-nCov pandemic began in Wuhan, China in December 2019 and
has caused extreme havoc in nearly the entire world. COVID-19, commonly
known as Coronavirus, is a novel highly contagious virus belonging to the
Coronaviridae family that was suspected of being transmitted from animals
to humans. This virus causes mild to serious respiratory disease and death.
This pandemic has engulfed 210 countries/regions in merely six months infecting
1,949,210 people and taking the death toll to 123,348. Presently, the highest cases
of 2019-nCoV infections have been reported in US, Spain, Italy and UK. However,
the cases are now abruptly rising in Brazil, Russia and India whereas China, the
place of origin of the disease, is now receiving a very few cases.

It is currently a global healthcare crisis worldwide with many countries' health care
systems falling apart. The only choice in this situation is to avoid the incidence of
infection and plan our healthcare system for the possible outcomes. Additionally,
ban on movement and lockdowns has caused a great impact on the growth and
financial aspects of economies.

In that reference, it is extremely crucial to create visualizations which help firms to

analyze this situation and to prepare themselves for the future. Additionally,
MedTourEasy, being one of the globally upcoming tele-medicine company in
global healthcare, it is important for the firm to understand the current situation of
COVID19 so as to gain more insights on the intensity of the pandemic, the
response of all countries and the impact it will have on their market. Also,
depending on the results of the analysis, this may be used for increasing their
market presence and capacity planning.
Hence, this project aims at collecting and analyzing large data sets to create intuitive
and interactive dashboards for representing COVID-19 in order to gain meaningful
insights. The project is majorly divided into 3 subsections, as below,

 Analysis of the problem: This is done to assess the gravity of COVID-19

worldwide and on India as well. It contains statistics and data representing
the problem – total cases, deaths, active and recovered. Also, it contains
many comparative statistics with respect to parameters like country, state,
age, gender etc.
  Analysis of the steps taken by the country: This is done to analyze how
various countries have reacted to this situation. It contains a thorough
analysis on different steps taken by different countries to mitigate the impact
of the pandemic on their country.

 A correlation between the above two sections to assess the impact of

strategies taken by countries on the spread of COVID-19

Each of the above sub-section has been represented in the form of dashboards which
are created using R language on RStudio IDE and RMarkdown package. These
dashboards use a wide array of functions and packages in R to create intuitive and
drillable dashboards, which can then be used by the firm to analyze the situation and
draw conclusions about the same.
1.3 Objectives and Deliverables
This project focuses on creating easily understandable, interactive and dynamic
dashboards by gathering data of COVID-19 from various sources like John Hopkins
Data Repository, ECDC, covid19india.org etc. and using the coding language R and
packages like readr, dplyr, ggplot, ggplot2, flex dashboard and other RShiny
Packages to visualize these statistics which will enable the firm to analyze the
situation and draw conclusions regarding the pandemic. The prototype for all the
dashboards will be created using Power BI (primarily to create dynamic
visualizations like world map, heat maps, forecasting, slicers etc.)

The project consists of 3 dashboards detailed as follows (3 deliverables):

a. Analysis of the problem of COVID19: This dashboard focuses on analyzing the

data regarding the problem of COVID19. It highlights the following points and
displays them through various types of visualizations:
 Country wise comparison of total cases – Total / Active / Recovered /
Deaths with a thorough analysis of cases increasing date wise.
 Country wise comparison of variation of cases over time. This will depict
how the cases have increased over each day.
 Comparison of top 10 countries with respect to increase in cases.
 Country wise comparison of variation of cases with respect to age group,
gender, health, ethnicity.
 Analysis of most vulnerable areas Worldwide.
 Analysis of impact of COVID-19 on GDP of a country and other factors

b. Analysis of steps taken by countries: This dashboard focuses on analyzing

various steps taken by the governments of many countries. It highlights the
following points (World level):
 Country wise comparison of number of tests done.
 Country wise comparison of number of tests and positives obtained over
time.
 Country wise comparison of number of total and new tests done per
million and thousand.
 Country wise comparison of lock down measures in terms of time of
implementation; duration; gravity of implementation and its impact of
COVID spread and deaths.
  Comparison of testing policies and travel controls adopted by different
countries.
 Comparison of total tests done with respect to GDP per capita of a
country.

c. Analysis of COVID-19 situation in India: This dashboard focuses primarily on

India’s situation in the pandemic, which highlights the following points:
 Overall and state wise comparison of Total Cases, Active, Recovered and
Deaths with a thorough analysis of cases increasing date wise.
 Analysis of most vulnerable areas in India.
 State wise comparison of variation of cases over time. This will depict
how the cases have increased over each day.
 Comparison of top 5 states with respect to increase in cases.
 Comparison of top 10 districts of the 5 majorly affected states.
 Comparison of cases with respect to age group, gender, diabetes
prevalence, hospitalizations etc.
 Overall comparison of Total tests conducted.
 Comparison of total tests per million and thousand, test positivity rate and
tests per confirmed cases with a thorough analysis of tests conducted date
wise.
 State wise comparison of number of tests done.
 State wise comparison of number of total and new tests done per million
and thousand and test positivity rate.
 I. METHODOLOGY
2.1 Flow of the Project
The project followed the following steps to accomplish the desired objectives and
deliverables. Each step has been explained in detail in the following section.

Gathering
Requirements
& Defining
Problem
Statement

Publishing Data
Dashboards Collection

Development
of Designing
Dashboards Databases

Prototypin Data
g Cleaning

Data
Filtering
2.2 Use Case Diagram

Above figure shows the use case of the project. There are two main actors in the same: The
Client and Developer. The developer will first gather requirements and define the problem
statement then collecting the required data and importing it. Then the developer will design
databases so as to identify various constraints and relations in the data. Next step is to clean
the data to remove irregular values, blank values etc. Next, exploratory data analysis is
conducted to filter the data according to the requirements of the project. Then a prototype
of the dashboards is created using PowerBI to get a clear view of the visualizations to be
developed. Finally, dashboard is developed and analyzed to publish the results to the client.
2.3 Language and Platform Used

2.3.1 Language: R

It is a programming language and software environment for statistical analysis,

representation of graphics, and reports. R was developed in the University of
Auckland, New Zealand by Ross Ihaka and Robert Gentleman, and is currently
being developed by the R Technology Core Team. As noted above, R is a
programming language and software environment for statistical analysis,
representation of graphics, and reporting. The important features of R are:
 R is a well-developed, simple, and effective programming language that
includes conditionals, loops, recursive functions defined by the user, and
input and output facilities.
 R has efficient data processing and storage facilities.
 R includes a set of operators for arrays, lists, vectors, and matrix calculations.
 R offers a detailed, coherent and organized data analysis tool set.
 R provides graphical data analysis facilities and displays either directly on
the computer or printing on papers.

2.3.2 IDE: RStudio

RStudio is an integrated development environment for R (IDE). It contains a

browser, syntax-highlighting editor supporting direct code execution, plotting,
history, debugging and workspace management tools. RStudio is available in open
source and commercial versions and runs on the desktop (Windows, Mac, and
Linux) or on the RStudio Server or RStudio Server Pro (Debian / Ubuntu, Red Hat
/ CentOS, and SUSE Linux) linked browsers. Major features are:
 RStudio runs on most desktops or on a server and accessed over the web.
 It integrates the tools you use with R into a single environment.
 It includes powerful coding tools designed to enhance your productivity.
 It enables rapid navigation to files and functions.
 It has integrated support for Git and Subversion.
 It supports authoring HTML, PDF, Word Documents, and slide shows.
 It supports interactive graphics with Shiny and ggvis.
2.3.3 Package: RMarkdown

R Markdown provides a data science authoring framework (.Rmd files). R

Markdown files can be used to save and execute code (also supports Python and
SQL), and produce high-quality reports that can be shared with an audience. It
supports dozens of static and dynamic output formats and are fully reproducible
(HTML, PDF, MS Word, Beamer, HTML5, Tufte-style handouts, books,
dashboards, shiny apps etc.)

2.3.4 Template: Flexdashboard

It is a template in RMarkdown files which is used to create a group of related

visualizations in the form of a dashboard. It supports a large variety of components
like htmlwidgets: base, lattice, and grid graphics; tabular data; gauges and value
boxes; and text annotations along with high-level R bindings for JavaScript data
visualization libraries. Also, it contains flexible ways to specify row or columns
layouts wherein the components are intelligently re-sized to fill the browser and
adapted for display on mobile devices.

2.3.5 Dynamic element: RShiny

Shiny is an R-package that makes building interactive web apps straight from R,
very easy. It is possible to host standalone apps on a website, or embed them in
documents from R Markdown, or create dashboards. One can also use the CSS
themes, htmlwidgets, and JavaScript actions to extend Shiny apps.

Installation:
install.packages(“rmarkdown”)
install.packages(“flexdashboard”)
runtime: shiny
Loading:
library(flexdashboard)
library(shiny)
This file contains the following parts:

 An (optional) YAML header surrounded by --- : The title, output format, orientation
and layout of the file are defined.
 R code chunks surrounded by ```{r} : It is in the arranged in the form of rows and
columns where the coding is to be done.
 Text mixed with simple text formatting. 
 Knit button: To run the file and display output.
 II. IMPLEMENTATION
3.1 Gathering Requirements and Defining Problem Statement
This is the first step wherein the requirements are collected from the clients to
understand the deliverables and goals to be achieved after which a problem
statement is defined which has to be adhered to while development of the project.

3.2 Data Collection and Importing

Data collection is a systematic approach for gathering and measuring information
from a variety of sources in order to obtain a complete and accurate picture of an
interest area. It helps an individual or organization to address specific questions,
determine outcomes and forecast future probabilities and patterns.

The data COVID-19 has been collected through various GitHub repositories,
mentioned as follows:

 Johns Hopkins University Center for Systems Science and Engineering

 European Centre for Disease Prevention and Control
 OurWorldData
 Covid19India.org

Data importing is referred to as uploading the required data into the coding
environment from internal sources (computer) or external sources (online websites
and data repositories). This data can then be manipulated, aggregated, filtered
according to the requirements and needs of the project.

Packages Used:

Readr: The goal of readr is to provide a fast and friendly way to read rectangular
data (like csv, tsv, and fwf). It is designed to flexibly parse many types of data
found in the wild, while still cleanly failing when data unexpectedly changes. To
accurately read a rectangular dataset with readr, one needs to combine two pieces:
a function that parses the overall file, and a column specification.

Readxl: The readxl package is used to get data out of Excel and into R. Compared
to many of the existing packages (e.g. gdata, xlsx, xlsReadWrite) readxl has no
external dependencies, so it’s easy to install and use on all operating systems. It is
 designed to work with tabular data. readxl supports both the legacy .xls format and
the modern xml-based .xlsx format.

Functions Used:

read.csv (): It is a wrapper function for read.table() that mandates a comma as

separator and uses the input file's first line as header that specifies the table's column
names. Thus, it is an ideal candidate to read CSV files. It has an additional
parameter of url() which is used to pull live data directly from GitHub repository.

read_excel (): It calls excel_format() to determine if path is xls or xlsx, based on

the file extension and the file itself, in that order.

Sample Code:

library(readxl)
library(readr)

x <- data.frame(read.csv(url(ECDC_ConfirmPath),stringsAsFactors = F))

vb <- data.frame(read_csv("D:/Traineeship/MedTourEasy - 6th April

2020/Project – COVID/Input Data/JHU/05-28-2020.csv"))

world1 <- data.frame(readxl::read_excel("D:/Traineeship/MedTourEasy - 6th

April 2020/Project - COVID/Input Data/ECDC/daily-cases-covid-
region.xlsx"))COVID/Input Data/JHU/05-28-2020.csv"))

3.3 Designing Databases

Once the data has been collected and imported into the R environment, it is
important to design the structure of the database tables so as to identify the
constraints in the data, keys, dependencies and relations between various tables.

Once the data is imported in the environment, it is converted into a data frame (data
type in R) which makes it easy to maintain the data in form of tables. The various
tables which have been created are mentioned as follows:

Attribute Data type Size Extra

Country_ID VARCHAR 5 Primary Key
Country_Name CHAR 15 Not Null, Unique
Latitude INT 3 Not Null
Longitude INT 3 Not Null
 Attribute Data type Size Extra
Country_ID VARCHAR 5 Foreign Key
Total Cases INT 15
Total Deaths CHAR 10
CFR INT 15
Median Age VARCHAR 10

Attribute Data type Size Extra

Country_ID VARCHAR 5 Foreign Key
GDP INT 15
Income Group CHAR 10
Total Cases INT 15
Total Deaths INT 15

Attribute Data type Size Extra

Country_ID VARCHAR 5 Foreign Key
Status CHAR 10 Not Null
Dates DATE -
Total Cases INT 15
New Cases INT 15
Total Cases / Mn INT 15
New Cases / Mn INT 15
Total Deaths INT 15
New Deaths INT 15
Total Deaths / INT 15
Mn
New Deaths / Mn INT 15

** Similar data frames are there for testing statistics and India statistics
3.4 Data Cleaning
“Quality data beats fancy algorithms”

Data is the most imperative aspect of Analytics and Machine Learning. Everywhere
in computing or business, data is required. But many a times, the data may be
incomplete, inconsistent or may contain missing values when it comes to the real
world. If the data is corrupted then the process may be impeded or inaccurate results
may be provided. Hence, Data cleaning is considered a foundational element of the
basic data science.

Data Cleaning means the process by which the incorrect, incomplete, inaccurate,
irrelevant or missing part of the data is identified and then modified, replaced or
deleted as needed.

With reference to the COVID-19 dataset, it may contain many null values or
incorrect value simply because of inconsistency in reporting cases and testing
statistics by countries and states. Hence various functions are used to clean this data.

Packages Used:

Tidyverse: It is a collection of essential data science R-packages. Under the

tidyverse umbrella, the packages help perform and interact with the data. There are
a whole host of things one can do with data, like sub setting, transforming,
visualizing and so on.

Dplyr: dplyr is a grammar of data manipulation, providing a consistent set of verbs

that help solve the most common data manipulation challenges. It is simply the
most useful package in R for data manipulation with the greatest advantage being
the use the pipe function “%>%” to combine different functions in R. From filtering
to grouping the data, this package does it all. It offers various functions like select,
filter, group_by, summarize etc.

Functions Used:

Is.na(): In R, missing values are represented by the symbol NA (not available).

Impossible values (e.g., dividing by zero) are represented by the symbol NaN (not
a number). This function is used to check if a dataset contains NA values or not.
 Na.rm(): When using a data frame function na.rm in r refers to the logical parameter
that tells the function whether or not to remove NA values from the calculation. It
literally means NA remove. It is a parameter used by several data frame functions.

Unique(): This function is used to filter out redundant data and keep only unique
values from the data frame.

Na.omit(): This function returns the object with listwise deletion of missing values.

Mutate(): This function adds new variables that are functions of existing variables

As.date(): This function is used to convert between character representations and

objects of class “Date” representing calendar dates.

Sample Code:

library(tidyverse)
library(dplyr)

cfr[is.na(cfr)] <- 0
cfr$Date <- anydate(cfr$Date)

xtot <- unique((x %>% filter(date >= as.Date('2020-02-15')))$location)

na.omit(s$Total.Confirmed)

3.5 Data Filtering

Data filtering is the method of choosing a smaller portion of the data set and using
that subset to view, analyze and evaluate data. Generally, filtering is temporary –
the entire data set is retained, but only part of it is used for calculation. It is also
called subsetting or drill down data wherein data is extracted with respect to certain
defined logical conditions. Filtering is used for the following tasks:

 Analyzing results for a particular period of time.

 Calculating results for particular groups of interest.
 Exclude erroneous or "bad" observations from an analysis.
 Train and validate statistical models.

With respect to COVID-19 dataset, the data needs to be filtered according to certain
conditions like dates between December and July, top 10 COVID-19 hit countries,
 gender, age groups etc.

Packages Used:

Tidyverse: It is a collection of essential data science R-packages. Under the

tidyverse umbrella, the packages help perform and interact with the data. There are
a whole host of things one can do with data, like sub setting, transforming,
visualizing and so on.

Dplyr: dplyr is a grammar of data manipulation, providing a consistent set of verbs

that help solve the most common data manipulation challenges. It is simply the
most useful package in R for data manipulation with the greatest advantage being
the use the pipe function “%>%” to combine different functions in R. From filtering
to grouping the data, this package does it all. It offers various functions like select,
filter, group_by, summarize etc.

Functions Used:
Slice(): This function is used to extract rows by position.
Filter(): This function is used to extract rows that meet a certain logical criteria.
Logical Comparisons:
<: for less than
>: for greater than
<=: for less than or equal to
>=: for greater than or equal to
==: for equal to each other
!=: not equal to each other
%in%: group membership. For example, “value %in% c(2, 3)” means that
value can takes 2 or 3.

Filter_all(), filter_at(): filter rows within a selection of variables. These functions

replicate the logical criteria over all variables or a selection of variables.
Sample_n(): This function randomly select n rows
Top_n(): This function selects top n rows ordered by a variable
Sample Code:

library(tidyverse)
library(dplyr)

x1 <- x %>% filter(date >= as.Date('2020-02-15'))

x2 <- x %>% filter(date == as.Date('2020-06-07'))
x2 <- x2 %>% top_n(10, total_cases_per_million)
Page 21 of
d <- x %>% filter(date == '2020-06-07')
46
3.6 Prototyping – Power BI
A prototype is an early version, model, or release of a product that is constructed to
test a design or process. It is generally used by system analysts and users to assess
a new design to enhance precision. Prototyping serves to specify a real, working
system rather than a theoretical one. Creation of a prototype in some design
workflow models is the step between formalizing and testing an idea.

Power BI is Microsoft's business analytics software. It aims to provide interactive

visualizations and business intelligence capabilities with an interface that is easy
enough to create your own reports and dashboards for end users. It provides cloud-
based BI services, known as "Power BI Services," along with the "Power BI
Desktop" desktop-based interface. It provides capabilities for data warehouse,
including data planning, data discovery and interactive dashboards. It has the
following features:

 Easy to connect, model, and visualize data, creating memorable reports

personalized with KPIs and brand.
 Can generate fast, AI-powered answers to business questions
 Data is better secured across Power BI reports, dashboards, and data sets with
persistent protection that keeps working even when shared outside the
organization or exported to other formats such as Excel, PowerPoint,
and PDF.
 Input: In the form of Excel, CSV, text, SQL and other formats
 Visualizations: Wide variety of graphs, infographics, KPIs, Filters, Slicers,
etc.
 Output: Easily publishable reports and dashboards

The screenshots of the prototype are as follows:

Page 22 of
46
Page 23 of
46
3.7 Development of Dashboards
As stated earlier, the dashboards have been created using the flexdashboard package
in R which provides a template in the form of rows or columns to display a group
of related data through interactive visualizations from which conclusions can be
drawn. The 3 dashboards to be developed are stated below:

 COVID-19 Worldwide Tracker

 COVID-19 Testing Tracker
 COVID-19 India Tracker
3.7.1 Defining Visuals

Data visualization is presenting data in a graphical or pictorial format. It allows

decision-makers to see visually presented analytics, so that they can grasp difficult
concepts or identify new patterns. In interactive visualizations, technology can be
used to dig in charts and graphs for more detail, interactively modifying what data
one can see and how it works.

Because of the way in which the human brain processes information, it is easier to
visualize large amounts of complex data using charts or graphs than to poring over
spreadsheets or reports. Data visualization is a quick, easy and universal way of
conveying concepts. Data visualization can also:

 Identify areas that need attention or improvement.

 Clarify which factors influence customer behaviour.
 Help you understand which products to place where.
 Predict sales volumes.

In R, these visualizations are based on the Grammar of Graphics. It is a tool that

enables one to concisely describe the components of a graphic. Such a grammar
allows us to move beyond named graphics (e.g., the ``scatterplot'') and gain insight
into the deep structure that underlies statistical graphics. It contains the following
layers:

1. Data: The data element is the data set itself. In this reference, the data is the
COVID-19 datasets.
2. Aesthetics: The data has to be mapped onto the aesthetics element (variables

Page 24 of
46
 mapped to x or y position and aesthetics attributes such as color, shape, or size)
3. Geometries: This element determines how the data is being displayed (bars,
points, lines). It consist of geom_line(), geom_scatter(), geom_bar(),
geom_col(), geom_area(), geom_point(), etc. Every single plot that is made will
always consist of the above three layers.
4. Facet: It is an optional layer. Facetting splits the data into subsets and displays
the same graph for every subset.
5. Statistics: It helps to transform the data (add mean, median, quartile)
6. Coordinates: It helps to transforms axes (changes spacing of displayed data)

Packages Used:

Ggplot2: Ggplot2 is a declarative graphics development framework focused on The

Grammar of Graphics. Once the user provides the data and tells ggplot2 how to map
aesthetic variables and what graphic primitives to use, it takes care of the details. In
most cases, one starts with ggplot(), supplies a dataset and aesthetic mapping
(with aes()), the adds on layers (like geom_point() or geom_histogram()), scales,
faceting specifications (like facet_wrap()) and coordinate systems
(like coord_flip()).

Plotly: This is a complement to the ggplot package which includes javascript

libraries to provide more interactive visuals.

Leaflet: Leaflet is one of the most popular open-source JavaScript libraries for
interactive maps. It makes it easy to integrate and control leaflet maps in R. Some
of its features include:
 Interactive panning/zooming
 Compose maps using arbitrary combinations of Map tiles, Markers,
Polygons, Lines, Popups, GeoJSON
 Create maps right from the R console or RStudio
 Embed maps in knitr/R Markdown documents and Shiny apps
 Easily render spatial objects from the sp or sf packages, or data frames with
latitude/longitude columns

Page 25 of
46
 Sample Code:

leaflet(map) %>%
addTiles() %>%
setView(lng = 0, lat = 30, zoom = 1.5) %>%
addPolygons(
fillOpacity = 0.7,
fillColor = ~ pal(Cases),
color = "white",
label = ~map_labels,
highlight = highlightOptions(
color = "black",
bringToFront = TRUE
)
)

a <- ggplot(xf, aes(date, new_cases, fill = location)) + geom_area(alpha=0.6)

+ labs(x="Date", y = "New Cases")

a <- a + theme_minimal() + theme(legend.position="bottom",

legend.title = element_blank(),
legend.background = element_rect(fill =
"oldlace", size = 0.5, linetype = "dash", colour = "black"),
axis.text.x = element_text(face="bold",
color="black", size=10, angle = 30, vjust=.5),
axis.text.y = element_text(face="bold",
color="black", size=10),
axis.line = element_line(colour = "black",
size = 1.5, linetype = "dashed")) + scale_y_continuous(labels = scales::comma)

ggplotly(a)

3.7.2 Integration with RShiny

As discussed, Shiny is an R-package that makes building interactive web apps

straight from R, very easy. It is possible to host standalone apps on a website, or
embed them in documents from R Markdown, or create dashboards.

Using Shiny, the visualizations can be made more interactive and drillable. This
means that the user, along with viewing the dashboards, can also give inputs to the
dashboard which will then automatically update its visuals. By adding
‘runtime:shiny’ to the title block, the document can be embedded with shiny app.

In reference to the project, a combination of plotly+shiny is used for making the

dashboard user interactive. Each shiny app consists of two major parts as follows:

1. The user interface, ui, describes how the web page displays inputs and output
widgets. The fluidPage() function offers a nice and quick way to get a grid-
Page 26 of
46
 based responsive layout, and the UI is completely customizable and packages
like shinydashboard make it easy to leverage more sophisticated layout
frameworks.

2. The server function, server, defines a mapping between input values and
output widgets. More precisely, the shiny server is a R function () between
client input values and Web server outputs.

Shiny comes with a handful of other useful pre-packages input widgets. Although
many shiny apps use them straight out of the box, CSS and/or SASS input widgets
can be easily stylized, and even customized input widgets can be integrated.

Some of the shiny widgets include:

 selectInput()/selectizeInput() for dropdown menus.

 numericInput() for a single number.
 sliderInput() for a numeric range.
 textInput() for a character string.
 dateInput() for a single date and dateRangeInput() for a range of dates.
 checkboxInput()/checkboxGroupInput()/radioButtons() for choosing a list
of options.

Sample Code:

ui <- fluidPage(
selectizeInput(
inputId = "worldregion",
label = NULL,
choices = c("Please choose a city" = "", worldregion),
multiple = TRUE
),
plotlyOutput(outputId = "p")
)

server <- function(input, output, session, ...) {

output$p <- renderPlotly({
req(input$worldregion)
if (identical(input$worldregion, "")) return(NULL)
a <- ggplot(data = filter(newerdata, Entity %in% input$worldregion)) +
geom_area(aes(Date, Cases, fill = Entity), alpha=0.9) + labs(x="Date", y
= "New Cases") + theme_minimal()
ggplotly(a, height=300, width = width) })}

shinyApp(ui, server)

Page 27 of
46
 III. SAMPLE SCREENSHOTS AND OBSERVATIONS
4.1 COVID-19 Worldwide Tracker (Data as on 14th June 2020)
4.1.1 Heading and Quick Facts

This is the topmost bar of the dashboard which shows the heading along with two
toggle buttons to represent the two pages of the dashboards – ‘World Situation’ and
‘Deep Dive’. 4 facts have been displayed as follows: Total Cases, Active Cases,
Deaths and Recovered. All the data is directly obtained from the GitHub repository
and is updated automatically.

4.1.2 World Hotspots

The first map of the dashboard is a heatmap of the world representing the total
confirmed cases in the world represented by the leaflet widget. On the right, there
is a legend which associates the case bins with the respective color. When a user
hovers over a particular country, more information is displayed as shown below.

Observations: The heatmap shows that the countries with the darkest colors have
been the worst affected. The top 5 worst hit countries include USA, Brazil, Russia,
India and UK.

Page 28 of
46
4.1.3 Day Wise Comparison - Total Cases and Deaths

The second visual is a bar chart representing the date wise increase in total cases
and deaths in the world. R Shiny has been used to make the chart user interactive.
The user can select any country and the chart will be updated according to it. Also,
onHover event has been added.

Observations: This visual indicates that there has been a linear increase in the total
number of cases and deaths worldwide with the total number of cases crossing 7.9
million and deaths crossing 0.4 million as on 14th June 2020.

4.1.4 Daily Confirmed Cases - By World Region

This visual represents the new cases emerging by date with respect to 6 world
regions. OnHover event has been added. Also, when a user double clicks on a
world region, that particular region is isolated on the graph.

Observations: This visual represents the variation of new cases in the major world
regions. In the initial months of February and March, the continents of Asia (China)
and Europe (Spain, Italy, UK) were faced an increasing trend in the number of new
cases everyday (indicated by larger area) whereas other continents were still int he
early stages of the pandemic. Towards the month of April, South America (Brazil)
and North America (USA) faced a peak in cases due to the countries reaching the
community stage of transmission. With regards to the last 2 months, there has been
a major peak in Asia again due to increasing number of cases in India and Russia,
SA and NA continue to have greater number of cases and European countries have
obtained a diminishing number of new cases per day.

Page 29 of
46
4.1.5 Top Countries - Total Confirmed Cases

This visual represents a line chart of the date wise increase in total cases by top
countries. R Shiny has been used to make the chart user interactive. The user can
select any country and the chart will be updated according to it. Also, onHover event
has been added.

Observations: This visual represents the variation of total cases by country. As

stated previously, China was the first country to face a peak in the total number of
cases which was soon surpassed by European nations like Spain, Italy and UK. Soon
after, due to less restrictions in movement, number of cases in USA increased
exponentially. In the second half of the pandemic, countries like India, Russia and
Brazil (which were able to delay the peak of total cases due to government measures
like lockdown and social distancing) are now facing a great rise in the cases
surpassing all European nations also.

Page 30 of
46
4.1.6 Day Wise Comparison - New Cases and Top 10 Countries by Total Cases per
Million

The first visual represents an area chart of the date wise increase in new cases by
countries and the second one represents a bar chart of the top 10 countries by total
cases per million. OnHover events are there on both and by selecting a country from
the legend, the statistics for that can be isolated.

Observations: The first visual represents the same observations as stated before.
The second visual represents the top 10 countries by total cases per million. It is
observed that if measured by the total cases per million, the top 10 countries are
entirely different to when measured by the total cases. This is because of the fact
that the smaller nations have a relatively smaller population which makes the total
cases per million higher indicating the fact that in countries like Qatar, Vatican,
Chile etc., it is easier to contain the spread of the virus due to lesser population and
restrictions of lockdown and social distancing can easily be implemented and
followed. This is the reason why these countries are not in the top 10 worst affected
list.

4.1.7 Cases and Deaths Vs. Income Group

These visuals represent a stacked area chart of date wise increase total cases and
deaths by income group. OnHover event and legend selection is also included.

Observations: The following graphs indicate the variation of cases by the income
group. As it is clearly highlighted, the high-income group has been the worst
affected, followed by the upper middle income, lower middle income and finally
the low-income group. Similar trend is highlighted in the variation of deaths as well.
The primary reason for the same is that initially, the virus spread to major countries
because of infected people travelling abroad and the major proportion of groups
who can afford air travel is from the high and upper middle-income groups

Page 31 of
46
 justifying the trends of the graph.

4.1.8 Total Cases Vs. GDP and Case Fatality Rate Vs. Median Age

The first visual represents a point chart of the GDP per Capita by Total cases and
Country whereas the second one shows median age by case fatality rate and
country. OnHover event and legend selection is also included.

Observations: These visuals indicate the countries with a greater GDP per Capita
have greater capacities to invest in healthcare facilities of their country. Also, the
countries which have a median age between 40 and 50 are facing greater number of
COVID cases.

4.1.9 Doubling Time

This visual represents a line chart of date wise variation in doubling rate by
country. OnHover event and legend selection is also included.

Observations: The data of this graph is as on 29th May 2020 after which the
reporting for doubling time has been inconsistent. This graph indicates that the
countries which faced a greater peak in total number of cases in the first stage of the
pandemic have been able to tackle the spread of virus. Countries like UK, Germany,
South Korea were able to do it ramping up their testing facilities whereas Spain,
Italy and France by imposing restrictions like lockdown, due to which the doubling

Page 32 of
46
 rate has increased to as high as 800 days. On the other hand, countries like Brazil,
India, Russia, Mexico etc. haven’t been able to increase the doubling rate due to an
exponential increase in the cases which can be regarded to lesser amount of testing
/ lesser contact tracing / relaxation of restrictions.

4.1.10 Diabetes Prevalence, Aged 70 Older and Hospital Beds per Thousand Vs. Total
Deaths

The following visuals show the variation of total deaths by parameters like
diabetes prevalence, aged 70 and hospital beds by country. OnHover event and
legend selection is also included.

Observations: These visuals depict a correlation between various factors which

affect the increase in total cases and deaths like diabetes prevalence, age and
presence of hospital beds.

Page 33 of
46
4.2 COVID-19 Testing Tracker (Data as on 14th June 2020)
4.2.1 Heading and Quick Facts

This is the topmost bar of the dashboard which shows the heading along with two
toggle buttons to represent the two pages of the dashboards – ‘Worldwide Situation’
and ‘Deep Dive’. 2 facts have been displayed as follows: Total Tests and Total Tests
per Thousand. All the data is directly obtained from the GitHub repository and is
updated automatically.

4.2.2 Total COVID-19 Tests and Tests per 1000 People

These visuals are line charts representing the date wise increase in total tests and
total tests per 1000 people by country. OnHover and legend selection event has been
added.

Observations: The first visual indicates how countries have ramped up their testing
facilities to contain the spread of the virus. Although, US has not reported the testing
conditions properly which is the reason for the broken line. In the first half of the
pandemic, countries like South Korea, Iran, Singapore, New Zealand etc. increased
their testing facilities rapidly which helped them to keep their cases in check, to the
extent that New Zealand declared itself COVID free recently. On the other hand,
insufficient testing in countries like Turkey, Italy, Spain, France etc. led to a
relatively greater number of cases. In the second half, India and Russia have
increased their testing and contact tracing facilities due to sudden increase in daily
new cases.

Page 34 of
46
 The second visual indicates that smaller countries like Denmark, Qatar, Australia
etc. have a greater test per 1000 people figure when compared to India, USA, Russia
even though the actual testing number of the latter countries is higher. This is
regarded to the fact that the former countries have a lesser population and hence
greater testing per 1000 people. These countries have been successfully able to
contain the spread of the virus.

4.2.3 Tests per 1000 since 100th Confirmed Case and 5th Death

These visuals are line charts representing the total tests per 1000 people after the
100th case and 5th death were detected by country. OnHover and legend selection
event has been added.

Observations: These graphs indicate how various countries have reacted to the
pandemic situation after reporting their 100th case and 5th death. It indicates very
contrasting situations throughout all the countries. As evident from the graph,
countries like Denmark, Qatar, New Zealand, Belgium, Australia etc. ramped up
their tests per 1000 exponentially just as soon as they reported the 100th case or the
5th death due to which they were able to keep the total number of cases in check. On
the other hand, countries like India, Peru and Mexico did not ramp up the testing as
required and hence the total number of cases are on the rise.

Page 35 of
46
4.2.4 COVID-19 Testing Policies and International Travel Controls

These maps of the dashboard represent a heatmap of the world by COVID-19 testing
policies followed and International travel controls imposed by various countries. It
is represented by the leaflet widget. On the right, there is a legend which associates
the bins with the respective color. When a user hovers over a particular country,
more information is displayed as shown below.

Observations: Countries which followed a more open testing policy have been able
to detect more cases and contain the spread. Also, major countries have imposed
complete ban on international travel within countries.

Page 36 of
46
4.2.5 Total tests vs. GDP per Capita

This visual represents a point chart of the GDP per Capita by Total tests and
Country. OnHover event and legend selection is also included.

Observations: This visual represents the variation of GDP per Capita by total tests
and countries. It is very evident that countries having a greater GDP per Capita are
able to invest more in their healthcare and testing facilities in contrast to those which
have a lower GDP per Capita.

4.2.6 Testing Status

This visual represents a line chart of the date wise increase in total deaths and tests
by each top country. R Shiny has been used to make the chart user interactive. The
user can select any country and the chart will be updated according to it. Also,
onHover event has been added.

Observations: This visual represents the date wise variation of total tests and deaths
by country. It indicates that the countries which were able to increase the tests
conducted by day, have been able to keep a check on the total deaths occurred due
to the fact that with more testing, cases can be detected faster and the government

Page 37 of
46
 can keep a check on the spread of the virus.

4.2.7 Lockdown Period

This visual represents a gantt chart of the start and end dates of the majorly
affected countries in the world. OnHover event has been added.

Observations: This graph represents a gantt chart of the dates of start and end of
lockdown of major countries. In case of USA, the lockdown was introduced at a
very later date which resulted in a magnanimous growth in number of cases. On the
other hand, New Zealand, Denmark, Belgium and Australia started the lockdown
just as soon as it reported the first case due to which the spread was contained. Also,
India started the lockdown when its 500th case was reported which helped it to delay
the peak cases for about one month. Additionally, South Korea and Singapore never
went into a complete lockdown but due to their excellent testing facilities, they were
able to keep the numbers in check.

Page 38 of
46
4.3 COVID-19 Testing Tracker (Data as on 14th June 2020)
4.3.1 Heading and Quick Facts

This is the topmost bar of the dashboard which shows the heading along with two
toggle buttons to represent the two pages of the dashboards – ‘Case Situation’ and
‘Testing Situation’. 4 facts have been displayed as follows: Total Cases, Deaths,
Recovered and Tests. All the data is directly obtained from the GitHub repository
and is updated automatically.

4.3.2 Confirmed Cases, Deaths and Recovered

These maps of the dashboard represent a heatmap of India by total cases, deaths and
recovered. It is represented by the leaflet widget. On the right, there is a legend
which associates the bins with the respective color. When a user hovers over a
particular country, more information is displayed as shown below.

Observations: As on 14th June 2020, total number of cases reported in India has
crossed the 3-lakh mark although the number of recoveries indicate a positive
picture wherein the recoveries are almost half that of the cases, pushing the recovery
rate to a 50% high mark. India has also managed to keep the number of deaths in
check by increased hospitalizations. The top 10 affected states are: Maharashtra,
Tamil Nadu, Delhi, Gujarat, UP, Rajasthan, West Bengal, Madhya Pradesh,
Haryana and Karnataka.

Page 39 of
46
4.3.3 Day Wise Comparison – Total and New Cases and Deaths

These visuals are line and column charts respectively representing the date wise
increase in total and new cases and deaths in the world. Also, onHover event has
been added.

Observations: These visuals indicate overall increase in total cases with time. The
total recovered cases have shown an increasing trend representing that India has
been able to provide adequate healthcare facilities to the infected patients. This is
also the reason why the total deaths have remained low when compared to other
countries.

Page 40 of
46
4.3.4 State Wise Comparison of New Confirmed Cases, Deaths and Recoveries

These visuals represent line charts of date wise variation of new cases, deaths and
recoveries by states of India. OnHover event is included.

Observations: These graphs show a drill down, state wise analysis of variation of
cases over time. The first case in India was reported in the state of Kerala on 30th
January. In the months of February and March, the spread of the virus was fairly
slower. India reported its 500th case on 25th March, when the country went into a
complete lockdown. Then after, the cases started rising majorly in states like
Maharashtra, Madhya Pradesh, Gujarat, Rajasthan and Delhi. As on 14th June,
Maharashtra has emerged to become the major hotspot in India (crossing 1 lakh
mark). On a positive note, the recoveries per state are also on a rise due to improved
healthcare facilities provided to its citizens.

Page 41 of
46
4.3.5 District Wise Comparison – Top 5 States

These visuals represent bar charts of district wise variation of total cases and deaths
of top 5 infected states of India. OnHover event is included.

Observations: These graphs represent the top 10 worst affected districts of the 5
majorly affected states of India indicating the fact that major metropolitan cities like
New Delhi, Mumbai, Chennai, Bangalore etc. have been reporting greater cases
when compared to smaller cities and districts.

Page 42 of
46
4.3.6 Demographics – Age and Gender

These visuals represent bar charts of gender and age variation of total cases of India.
OnHover event is included.

Observations: These graphs represent the overall demographics of the cases

indicating that the ratio of infected males is greater and people between the age
group of 21 and 30 are most infected.

4.3.7 Total COVID-19 Tests and Tests per Million

These visuals represent line charts of date wise variation of total tests and tests per
million conducted in India. OnHover event is included.

Observations: The first visual indicates how India has ramped up its testing
facilities to contain the spread of the virus. Although, the tests per million still
remains low due to a greater population.

4.3.8 State Wise COVID-19 Tests, Tests per 1000 People, Tests per Million

These visuals represent stacked area charts of date wise variation of total tests, tests
per 1000 and tests per million by states of India. OnHover event and legend selection
is included.

Observations: These visuals indicate the state wise distribution of tests reported
Page 43 of
46
every day. Kerala was successfully able to contain the spread, even though it
reported the first case in India, can be attributed to the fact that it followed policies
of extensive contact tracing and testing. A similar pattern was followed by Rajasthan
and Karnataka which reported very high number of cases initially but were able to
bring to down the numbers in later stages of the pandemic. In Maharashtra, there
has been an exponential rise in cases without signs of slowing down which can be
attributed to the fact that Mumbai emerged as a hotspot due to crammed living
conditions and inconsistent testing and contact tracing. In contrast, states like Delhi,
Gujarat and Tamil Nadu, even though with excellent test reporting, have been facing
greater number of cases daily.

Page 44 of
46
 IV. CONCLUSION AND FUTURE SCOPE
The entire world is in midst of a serious pandemic which has affected more than 200
countries causing more than 7 million infected cases and 0.4 million deaths. This pandemic
has taken its economic and financial toll on most of the major economies of the world.

This project aimed at analyzing the current situation of the pandemic by creating intuitive
and user interactive dashboards and drawing conclusions on the impact it will have on the
world. Currently, the project is in its last stage of development with the dashboards been
developed and submitted for review and feedback.

With regards to the future work, the firm aims at regularly updating the dashboards with
time and integrating it with their systems so as to continually draw conclusions and analyze
the results. This will enable them to predict future business opportunities and provide a
basis on which they can plan on increasing their market presence and capacity planning.

Page 45 of
46
 V. REFERENCES
Data Collection

The following websites have been referred to obtain the input data and statistics:

a. https://api.covid19india.org/

b. https://www.aa.com.tr/en/latest-on-coronavirus-outbreak/worldwide-covid-19-
testing-ratio-per-country-million/1800124#

c. https://www.tableau.com/covid-19-coronavirus-data-resources

d. https://github.com/CSSEGISandData/COVID-19/tree/master/csse_covid_19_data

e. https://github.com/owid/covid-19-data/tree/master/public/data

Programming References

The following websites have been referred for R coding and Shiny tutorials:

a. https://datascienceplus.com/category/programming

b. https://rstudio.com/resources/webinars/

c. https://bookdown.org/yihui/rmarkdown/document-templates.html

d. https://datascienceplus.com/map-visualization-of-covid19-across-world/

e. https://bookdown.org/yihui/rmarkdown/dashboards.html

f. https://rmarkdown.rstudio.com/lesson-12.html

g. https://bookdown.org/yihui/rmarkdown/cheat-sheets.html

h. http://www.htmlwidgets.org/showcase_leaflet.html

i. http://jeffgoldsmith.com/p8105_f2017/shiny.html

j. https://rmarkdown.rstudio.com/flexdashboard/using.html#page_icons

Page 46 of
46