Flu And Coronavirus Simulator (FACS)¶
FACS is an agent-based modelling code that models the spread of flu and coronaviruses in local regions. Up to now, we have used it to model the spread of Covid-19 in various regions of Europe:
- The United Kingdom: Greater London, South-East and North-West of England
- Lithuania: The region of Klaipėda
- Romania: The city of Călărași
- Türkiye: The cities of Ankara and Istanbul
The code can be repurposed to model other regions, and its current (sequential) implementation should be able to run up to 2 million households within a reasonable time frame. It also supports the implementation of lockdown measures, vaccination programmes, track and trace and mutated versions of the virus.
What sets FACS apart from many other codes is that we have a partially automated location extraction approach from OpenStreetMaps data (the scripts reside at https://www.github.com/djgroen/covid19-preprocess), that we resolve a wide range of different location types (e.g., supermarkets, offices, parks, schools, leisure locations and hospitals) and that we have a specific algorithm for modeling infections within these locations, taking into account the physical size of each location.
We are currently finalizing a first journal paper about FACS and will link to it from this page once it has become available.
Quick installation notes¶
To install FACS,
- Clone the GitHub repository
git clone https://github.com/djgroen/facs.git
- Change to the newly created
facsdirectory
cd facs
- Running requirements.txt will set up Python-required dependencies, however, you must first carry out the steps listed below if your computer does not already have a Meggase Passing Interface (MPI) installation.
Note: In the next step, since we are installing software, we would need admin or root privileges.
We have provided instructions for installing MPICH (a variant of MPI) on Ubuntu, however, for other Linux distributions, please refer to MPI user guide and manual.
sudo apt-get install python3 python3-dev python3-pip
sudo apt-get install mpich
pip install --no-cache-dir mpi4py
- With MPI and mpi4py (MPI for Python) installed, we can continue with installing dependencies:
pip install -r requirements.txt
How to run the code¶
To run a simple simulation of a basic test dataset, type:
python3 run.py --location=test
To run a simulation of the Borough of Brent, type:
python3 run.py --location=brent
Note that in this case, we explicitly specify the measures YML file that we wish to use. This is done in the format: covid_data/<measures_yml>.yml. So in this case we want to use the file covid_data/measures.yml as the measures file for Brent, we type:
python3 run.py --location=brent --measures_yml=measures_uk
Outputs are written as CSV files in the output_dir. E.g. for the test run you will get: covid_out_infections.csv out.csv
We also included a simple plotting script with the following syntax:
python3 PlotSEIR.py <output_filename> <plot_filename>
For example, if we wanted to plot the results from the previous FACS run for brent, we would type:
python3 PlotSEIR.py brent-measures.csv brent-plot
This will create a file called brent-plot.html in the current directory. For more detailed plots, we use FabCovid19.
Upadating FACS¶
The FACS code is currently under active development. In order to update FACS to the latest version, from the facs directory, run the command
git pull
Citing FACS¶
Our journal paper on FACS has now been accepted, and can be found here: https://bura.brunel.ac.uk/handle/2438/20914
The BibTex code is here:
@article{mahmood2020facs,
title={FACS: A geospatial agent-based simulator for analyzing COVID-19 spread and public health measures on local regions},
author={Mahmood, Imran and Arabnejad, Hamid and Suleimenova, Diana and Sassoon, Isabel and Marshan, Alaa and Serrano, Alan and Louvieris, Panos and Anagnostou, Anastasia and Taylor, S and Bell, David and others},
year={2020},
journal={Journal of Simulation (in press)}
publisher={Taylor \& Francis}
}