README.md

SEIR_Simple_SMA

About the projet

The project is a simulation made from scratch of a multi-agent based simulation of a SEIR/SEIRS infection model. This code is aimed to be as repeatable and reproducible as possible and is part of a collaborative work for a workshop from the RED network.

Agents are autonomous, walks randomly in a squared environment.

You can chance the parameters of the simulation in the parameters.yaml file (see the setting the parameters section for more information).

Built With

Intelij IDEA, Maven

Getting started

Prerequisites

Java 17+, Python 3.10

Installation

1. Clone the repo

git clone git@github.com:AxelCarayon/SEIR_Simple_SMA.git

2. Install dependencies

mvn clean install

3. Run simulation alone

java -jar out/artifacts/SEIR_Simple_SMA_jar/SEIR_Simple_SMA.jar

3. Run from python script and generate stats

python3 outputToGraph.py

Setting the parameters

On start, the simulation get the parameters from the file parameters.yaml. Below is a list of each parameters and what they means.

graphicalMode : Boolean

When true, will display an IHM when running a simulation

incubationRate : Float [0..1]

Chances each cycle that an EXPOSED agent become INFECTED.

infectionRate : Float [0..1]

Chances each cycle that an SUCEPTIBLE agent become EXPOSED if an INFECTED agent is nearby.

looseImmunityRate : Float [0..1]

Chances each cycle that an RECOVERED agent become SUCEPTIBLE again.

recoveryRate : Float [0..1]

Chances each cycle that an INFECTED agent become RECOVERED.

nbOfCycles : Integer

Determines for how much cycle the simulation will run.

Setting it to -1 or lower will run the simulation until it's manually stopped.

nbOfPatientZero : Integer

Determines how much agents will be already INFECTED when the simulation starts.

population : Integer

How much agents will be present in the simuation.

recordExperiment : Boolean

If true, the order in wich the agents wake up in the simuation will be recorded.

playRecord : Boolean

If true, will use a scheduler that will wake up agents accordingly to the record.

seed : Integer

Seed all the random elements in the simulation.

See the How the random works section for more details.

size : Integer

Set the size in pixels of the square world.

wrappingWorld : Boolean

If true, agents will be allowed to wrap to the other side when close to a border.

synchronousMode : Boolean