This github page contains the code and input data for the interactive COVID-19 mapping tool developed by the Vaccine Centre at the London School of Hygiene and Tropical Medicine.
Input data are obtained from the Johns Hopkins Center for Systems Science and Engineering github page.
The Shiny app, first launched on 11th February 2020, aims to complement existing COVID-19 mapping dashboards (such as those developed by the WHO and Johns Hopkins University) with several interactive features, including the timeline function and the ability to overlay past outbreaks.
Follow this link for the interactive Shiny app. A screenshot of the interface is provided below.
Key elements of the analysis code are as follows:
- jhu_data_full.R – an R script that extracts and reformats time-series from the Johns Hopkins Center for Systems Science and Engineering github page. The output files are saved in the input_data folder.
- ny_data_us.R – an R script that extracts and reformats time-series from the New York Times github page. The output files are saved in the input_data folder.
- app.R - an R script used to render the Shiny app. This consists of several plotting functions as well as the ui (user interface) and server code required to render the Shiny app. The script has become more complex over time as a growing number of interactive features has been added.
- input_data - a folder containing dynamic input data relating to the evolving COVID-19 pandemic (updated by jhu_data_full.R and ny_data_us.R) and static input data relating to past epidemics and country mapping coordinates.
From 1st May 2020 onwards, this github page will no longer be updated daily. The Shiny app automatically updates itself based on the code in jhu_data_full.R and ny_data_us.R and updated case data can be downloaded directly from the app via the 'Data' tab. To create up-to-date versions of the input data, users can clone this repository and run jhu_data_full.R and ny_data_us.R locally. Updated input data will also be uploaded to this github page at regular intervals.
From 3rd September onwards, the numbers of recovered cases are no longer plotted owing to large discrepancies in reporting rates among countries.
Several resources proved invaluable when building this app, including:
- A tutorial by Florianne Verkroost on building interactive maps;
- The SuperZIP app and associated code;
- The RStudio Leaflet tutorials.
Dr Edward Parker, The Vaccine Centre, London School of Hygiene & Tropical Medicine
Quentin Leclerc, Department of Infectious Disease Epidemiology, London School of Hygiene & Tropical Medicine