The model which I intend to build will simulate the spread of a virus through a community. The community will consist of a user-specified population of agents who move and encounter other agents in a stochastic manner. Initially, there will be a user-defined population of ‘infected’ agents and at each time step, a randomly selected cohort of agents will move and encounter other agents. The number of agents which move on each iteration is bounded by two parameters and a random value in this range is selected on each iteration. If an agent comes into contact with an agent who is infected (i.e. the agents are within a certain collision distance; another user-specified parameter), they themselves will become infected for a period of time. Currently, this period is set to 14 time steps but this is another parameter of the model which can be altered for experimental purposes.

There are a number of probabilities which will be associated with the model:

• Susceptible probability: The probability that an agent is susceptible to the virus. Upon initialisation, some agents will be immune to the virus.
• Immunity probability: The probability that an agent becomes immune to the virus after having been infected. In general, once infected, most agents become immune to a virus once they recover. However, this is not always the case.
• Death probability: The probability that an agent who becomes infected dies from the virus.

Again, these probabilities are defined by the user so different configurations of these probabilities can be tested.

To run a simulation, follow the below steps:

1. Alter the parameters if desired in the abm-virus.py file. The CSV files which produce the time-based graphs of output results are produced after each simulation. The directory in which these CSV files are stored can be altered using the path_out variable in the abm-virus.py file.

2. Simply run:

python abm-virus.py

while in the current directory.