Skip to content

This is the COVID-19 CovidSim microsimulation model developed by the MRC Centre for Global Infectious Disease Analysis hosted at Imperial College, London.


Notifications You must be signed in to change notification settings


Repository files navigation

COVID-19 CovidSim Model

This is the COVID-19 CovidSim microsimulation model developed by the MRC Centre for Global Infectious Disease Analysis hosted at Imperial College, London.

CovidSim models the transmission dynamics and severity of COVID-19 infections throughout a spatially and socially structured population over time. It enables modelling of how intervention policies and healthcare provision affect the spread of COVID-19. It is used to inform health policy by making quantitative forecasts of (for example) cases, deaths and hospitalisations, and how these will vary depending on which specific interventions, such as social distancing, are enacted.

With parameter changes, it can be used to model other respiratory viruses, such as influenza.


⚠️ This code is released with no support.

⚠️ This model is in active development and so parameter name and behaviours, and output file formats will change without notice.

⚠️ The model is stochastic. Multiple runs with different seeds should be undertaken to see average behaviour. This can now be done easily with the /NR command line parameter. The model code behaves deterministically if run with the same number of threads enabled and run with the same random number seeds.

⚠️ As with any mathematical model, it is easy to misconfigure inputs and therefore get meaningless outputs. The Imperial College COVID-19 team only endorses outputs it has itself generated.


Action Status Action Status Action Status Action Status Action Status

This model is in active development and subject to significant code changes to:

  • Enable modelling of more geographies

  • Enable modelling of different intervention scenarios

  • Improve performance


The model is written in C++.

The primary platforms it has been developed and tested on are Windows and Ubuntu Linux.

It should build and run on any other POSIX compliant Unix-like platform (for example macOS, other Linux distributions). However, no active development occurs on them.

Running the model for the whole of the UK requires approximately 20GB of RAM. Other regions will require different amounts of memory (some up to 256GB).

It is strongly recommended to build the model with OpenMP support enabled to improve performance on multi-core processors. 24 to 32 core Xeon systems give optimal performance for large (e.g. UK, US) populations.

See for detailed build instructions.


From within your build directory do:

make test
# If you want more progress indication
make test ARGS="-V"
# To parallelise tests add a -jN option for instance:
make test ARGS="-j6"
make test ARGS="-j6 -V"
# or
ctest -V
ctest -V -j6
# etc...

IMPORTANT: The test scripts use test data only are not runs reflective of real-world situations.

Sample Data

The directory data contains sample data.

The Python script demonstrates how to invoke CovidSim to use this data. See the sample README for details on how to run the samples.

The directory report9 contains files to allow the results tables in the Imperial College Report 9 - Impact of non-pharmaceutical interventions (NPIs) to reduce COVID-19 mortality and healthcare demand to be reproduced.


Model documentation can be found in the docs directory. Of particular interest are:

Given the entire Imperial College team is working full-time on the COVID-19 response, documentation is currently sparse. More documentation and sample files will be added as time permits. In the coming few weeks this will include a much more extensive set of input files to model strategies for exiting lockdown.

Relevant papers

The following papers are relevant to the model. Please note that some of them may require a subscription.

Copyright and Licensing

The source code for CovidSim is licensed under the GPLv3, see

It is Copyright Imperial College of Science, Technology and Medicine. The lead developers are Neil Ferguson, Gemma Nedjati-Gilani and Daniel Laydon.

Additional contributions for open-sourcing made by Imperial College of Science, Technology and Medicine, GitHub Inc, and John Carmack are copyright the authors.

Licensing details for material from other projects may be found in In summary:

CovidSim includes code modified from RANLIB which is licensed under the LGPLv3.

Sample data in the repository has been derived from the following sources:

WorldPop ( - School of Geography and Environmental Science, University of Southampton; Department of Geography and Geosciences, University of Louisville; Departement de Geographie, Universite de Namur) and Center for International Earth Science Information Network (CIESIN), Columbia University (2018). Global High Resolution Population Denominators Project - Funded by The Bill and Melinda Gates Foundation (OPP1134076).

WorldPop is licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0). The text of the license can be found at:


Due to time pressure on the development team, we are unable to provide user support at this time.

If you find issues with the code please raise them in our Issue Tracker.

This repository has a code of conduct which is detailed in the code of conduct. When raising an issue in this repository you agree to abide by the code of conduct.


This is the COVID-19 CovidSim microsimulation model developed by the MRC Centre for Global Infectious Disease Analysis hosted at Imperial College, London.



Code of conduct