Introduction

The Novel COVID-19 Coronavirus is the hottest topic right now. Every day, the media and newspapers share the number of new cases and deaths in several countries, try to measure the impacts of the virus on citizens and remind us to stay home in order to stay safe. The Coronavirus is on everyone’s lips.

In addition to governments, media and companies discussing about it, data scientists and data professionals in general are putting their knowledge and time at the service of the virus. This leads to a proliferation of applications, dashboards, blog posts, videos, datasets and code analyzing, in one way or another, the expansion of the COVID-19 and how it spreads in the population.

Top R resources on Coronavirus

As a data lover myself, I discovered a multitude of great resources about the Coronavirus. However, these resources were spread all over the internet and were often hidden by the mass of information of another type (e.g., alarming headlines, names of infected celebrities, companies showing off how they helped health care agencies, etc.). To tackle this issue, I collected and then shared in a previous article the best R resources on the Coronavirus I came across.

Note that only resources on R are shared in this article as R is my favorite statistical program and the one I am most familiar with. The fact that I use this program almost daily makes it easier for me to realize the complexity and time put behind a resource, and appreciate its quality and its potential. I am sure that there are other very interesting resources online (see for example the probably most popular dashboard by the Johns Hopkins Coronavirus Resource Center).¹ Nonetheless, a lot of people are in a better position than I am to judge the quality of resources made with programming languages for which I am not competent.

This article made me discover so many great resources about the Coronavirus and I keep receiving data visualizations and data analyses from scientists all over the world so that I include them in the collection. Thanks for that, it continuously improves the quality and completeness of the collection.

In addition to receiving R resources, a question often asked by readers was “How can I create a dashboard myself?” or “How can I build a dashboard specific to my country?”. I therefore thought it would serve some people if I created a dashboard specific to my country (Belgium) and detailed the steps on how to build it.

Questions on how to develop such dashboards came mostly from R beginners as advanced R users most probably know how to do one, or at least can easily use the resources I collected here as sources of inspiration for their own work. Furthermore, in response to the craze about the Coronavirus, interested users were quite in a hurry and wanted to have their own dashboard running as quickly as possible.

These questions led me to the idea of creating a simple (yet powerful and visually appealing) dashboard, as opposed to a Shiny app. Shiny apps have the advantage that they are interactive in the sense that users can edit the outputs and the visualizations by simply changing some inputs in a user-friendly way, while dashboards are static and cannot be modified by the final user. On the other hand, the advantage of a dashboard over a Shiny app is that it is much easier to code, especially if you are already proficient in R Markdown.

Coronavirus dashboard: the case of Belgium

Amongst all the visualizations I have seen so far, one is standing out by its simplicity and at the same time, by its completeness and by the quality of its visualizations. I thus decided to create a Coronavirus dashboard based on this already existing dashboard by Rami Krispin (which comes with a license that allows to be freely adapted and shared) and adapt it so that it is specific to Belgium. Note that I also removed some visualizations and tables from the initial dashboard to keep it really simple and straight to the point.

Before reading further, here is my Coronavirus dashboard adapted to Belgium and previews of the main sections below:

The dashboard is segmented into several sections that can be selected at the top:

• The summary section provides key measures about the Coronavirus (total cases, active cases and deaths) and a plot displaying the cumulative number of active cases and deaths from January 22, 2020 to the latest available date.
• The comparison section presents a comparison of the number of daily new cases (left panel) and the distribution of cases by type (right panel) with other European countries (you can also change these countries by replacing them in the code).
• The map section shows a world map of the confirmed cases, recovered cases and deaths. You can uncheck one or several types of cases (top right corner) and zoom in or out (top left corner) to adapt the map to your needs.
• The about section gives more information about the data, the dashboard in general and how often it is updated and some contact details in case someone finds a bug or a mistake.

I believe this simple dashboard is easy enough to be adapted to any country (and by anyone from beginner to expert), and still communicate key measures about the virus throughout some visualizations. A little extra which is worth mentioning is the fact that all plots are generated with the {plotly} package. This package allows to enhance plots by displaying additional relevant information when hovering over them (try by yourself!).

How to create your own Coronavirus dashboard

If you want to build your own dashboard specific to a country, follow these steps:

1. Open the dashboard here
2. See the entire code via the button “Source code” located in the top right corner of the dashboard, or see the code on GitHub. Copy that code.
3. Open a new R Markdown file (.Rmd), type any title and author (they will be replaced in the next step anyway), select HTML as the output format and click on OK:

4. Remove all the template code already present and paste the code you copied in step 1.
5. Make sure that the required packages are installed:

install.packages(c("devtools", "flexdashboard"))
devtools::install_github("RamiKrispin/coronavirus")

6. In the code, replace Belgium with your country. Change also the title and the author at the top of the document, and edit the about section at the bottom of the document.
7. Knit the document (see this article if you are unfamiliar with R Markdown). Your dashboard should appear in HTML format.

Following these 7 steps, you should already have a simple dashboard specific to your country. I have intentionally kept it simple so that everyone could copy it and have their own dashboard in a limited amount of time.

If you are familiar with the {flexdashboard}, {plotly} and {leaflet} packages for the dashboard interface and the visualizations, and the {dplyr} and {tidyr} packages for the data manipulation, feel free to edit the code according to your needs and improve your dashboard.

Additional notes

install.packages("devtools")
devtools::install_github("RamiKrispin/coronavirus")