COVID-19 vaccination

Code repository for the research article Relaxing restrictions at the pace of vaccination increases freedom and guards against further COVID-19 waves in Europe, published as a pre-print on arXiv. The data generation code is written in the Rust programming language, the plots are generated using Matplotlib in Python. To be able to run the scripts to reproduce the figures from the article, you need to have installed both Rust and Python and run the Makefile provided in this repository to first compile the Rust code on your machine.

The structure of this repository:

In this repository you will find five main folders:

• the scripts folder: In this you find three scripts to reproduce the main figures from the paper (see below).
• the figures folder, into which the resulting figures from the scripts get saved. It already
• the ICU_durations and the scenarios folders, in which the Rust and Python code lie, which is used to generate the figures and get called from the two scripts in the scripts folder.
• the model_lib folder: Here are the library files for our implementation of the SEIR+ICU+vaccination model from our paper. Look into the source code here if you look for implementation details.

To reproduce the figures from the paper:

All the scripts allow to change parameters and are set in a default state which already reproduces most of the main figures from the paper. For some parts of the figures, it is necessary to change some of the parameters in the scripts. Example figures for all three scripts with default parameters can already be found in the scripts folder. When changing any of the parameters in any of the scripts, please always make sure that the number of decimal places you type in matches the number of decimal places of the default values. Otherwise, the code will not be able to find the data files and the generation of the figures will fail. The correct number of decimal places is also always indicated in a comment following each parameter.

• To reproduce the scenarios featured in Figs. 1-3 and in the supplementary figures in our paper, there is a script scenarios.sh in the scripts folder. Running this will generate the data for all the considered scenarios and save two overview figures in figures/scenarios/:

  • One file named "scenarios_eta#.##uptake#.##(country code)_(contact structure).pdf", where the "#.##" represent the chosen values for the vaccine efficacy against infection and the final vaccine uptake respectively. This figure presents an overview of the dynamics of the most important parameters and compartments of all scenarios.
  • One file similarily named "scenarios_summary_eta#.##uptake#.##(country code)_(contact structure).pdf". This file includes four summary plots of the scenarios: A combined plot of all the gross reproduction number (Rt) profiles and ICU occupancy (as in e.g. Figure 1 of the main paper), as well as bar plots summarizing the total deaths and cases until the end of the vaccination period (independent on the given final uptake this is chosen to always be set to when an uptake of 80% is reached). The script also allows changing some default model parameters, like the vaccine efficacy and the total vaccine uptake as well as parameters characterising the scenarios, like the maximal gross reproduction number to which the restrictions are lifted in the end. Changing e.g. vaccine uptake from 0.8 to 0.9 and 0.7 yields the necessary plots for Figure 2 J,K.

• To reproduce the plots from Fig. 4 and 5 in the paper, there is another script ICU_durations.sh in the scripts folder. This will create a figure in figures/ICU_durations/, where the months that ICUs will have to work at full capacity given a lifting of the COVID-19 restrictions after the vaccination programs is plotted. Generating the data will take some time. This script also allows to change some parameters, like the maximal gross reproduction number to which the restrictions are lifted in the end (e.g. for Figure 4 B this is R_max=3.5, for Figure 4 C R_max=2.5). It also allows changing the demographics to match a given country (between GER, FIN, ITA and CZE) and the contact structure (between homogeneous contacts, pre-COVID contacts and pre-COVID contacts with reduced contagious contacts in schools). This allows to reproduce Fig. 5 and the supplementary Figures. NOTE: For some parameter combinations there sometimes appear some false (!) data points, i.e. data points where an ICU duration of 0 is computed from the program even though the uptake is much too low to prevent another wave. That is a known bug in the code. However, they are generally easy to spot.

• To reproduce the sensitivity analysis results from the Supplementary Information S2, there is a third script sensitivity.sh. This will generate multiple pdf files in figures/sensitivity/ (appropriately named after the parameter that is swiped through), which connect to the rows of the sensitivity analysis Figure S1. It also allows to change the default values, and which values for the parameters are swiped through.

How to run the scripts:

To run the scripts you need to have installed both Rust and Python. In the parent directroy (where this README is located) first run

make

to compile the Rust code. Then you can run the scripts individually with:

./scripts/scenarios.sh

./scripts/sensitivity.sh

./scripts/ICU_durations.sh