Skip to content

An agent-based model showing emergent behaviour from transport inequality.

License

Notifications You must be signed in to change notification settings

cohenrobinson/commute_model

Repository files navigation

commute_model

Commute Model is an Agent-Based Model (ABM) set at showing emergent behaviour of transport inequality. Currently when run in default, the agents of wealth form a high dense neighbourhood in the centre whereas the agents in poverty are sparce in the outskirts. This not only shows emergent behaviour that is common to today's societies, but leaves room to impliment a solution within the model.

Model Characteristics

The model starts all agents with some initial wealth and randomly distrbutes them around the city. Further distances from the city centre have less avaliability of public transport (PT). Car commuting has a cost per unit distance the agent has to travel to get to the city.

Agent Behaviour

Agents have the following behaviour:

  • A cost of living,
  • The ability to commute to work,
  • Earn an income from working,
  • Can choose to commute either by car or PT,
  • If an agent gains enough wealth, then it can move closer to the city.

Using the model

To run the model, execute 'run.py' with python. This will open up a local mesa server with the ABM running in this model.

Reading the model

On the grid, each agent is represented by a circle; black is the city centre; green is a wealthy agent; yellow is an average agent; red is a poor agent.

The charts below the model show the Gini coefficent of the model over time and the second chart is average distance at where the agents of a particular wealth bracket are located over time.