Agent-based Modeling of COVID19 using age distribution and family structure

Code for the preprint: The Role of Age Distribution and Family Structure on COVID-19 Dynamics: A Preliminary Modeling Assessment for Hubei and Lombardy

Running experiments

• To get started: python3 run_simulation.py

• To run simulations of different age-based policies python3 simulate_agepolicies.py -h

• To run large scale parameter sweeps python3 parameter_sweep.py -h

  • Also see: README_parameter_sweep.txt

The simulator

See seir_individual.py

Fitting comorbidity models

See comorbidity_inference.py. Estimates from the paper are already saved, so you can start doing simulations without rerunning this step unless you would like to fit your own model.

Data sources needed to run for specific country

Data already available in model (most countries)

1. Age distribution: Source
2. Age specific fertility of mothers: Source
3. Contact Matrices: Source

Outside data needed:

4. Distribution of household types: single household, couple without children, single parent +1/2/3 children, couple +1/2/3 children, family without a nucleus, nucleus with other persons, households with two or more nuclei (a and b)
5. Age specific rates of diabetes
6. Age specific rates of hypertension