Reconstructing the global dynamics of unreported COVID-19 cases and infections

Using reported data on COVID-19 cases and fatalities globally, we estimated the proportion of symptomatic cases that were reported in 210 countries and territories. We then use these estimates to attempt to reconstruct the pandemic.

Contents of the repository

The repo contains all the functions and scripts required to reproduce the results of the paper: Reconstructing the early global dynamics of under-ascertained COVID-19 cases and infections; Russell et al. (2020). Specifically, the functions and scripts in this repo are for:

• downloading the required data from the ECDC here
• importing the estimates from the bayesian inference model which forms the basis for the paper, the code and estimates for which can be found here
• combining the data and estimates to adjust the case curves
• producing all figures in the manuscript

To download the code, clone this repository using the command

git clone https://github.com/thimotei/covid_underreporting

To reproduce all of the figures in the manuscript, run the script

scripts/plotting_main_figures.R

in which all of the data required for the figures is pulled using the R/data_helper_functions.R and the plots are made using both the R/plotting_functions.R and R/plotting_helper_functions.R.

Finally, to reproduce all of the results quoted in the main text of the manuscript, run the script

scripts/miscellaneous_results.R