Overview

  • This program is a stochastic agent based simulation model of a towns local free houses. Focused on the town of St. Andrews. A 3D generated map of the environment is created to scale by pulling from geolocation data available from the Google Maps API.

  • The program is intended to be outsourced to entrepreneurs looking to open a pub in a predecided area and view the causal impact on the local economy. As we dream of a Wetherspoons opening in St. Andrews, the best way to argue for its worth is by simulating a Friday night in KY16 with and without a spoons modelled.

Tech Stack

  • Hand-rolled stochastic model in Python for the backend modelling
  • Python web sockets to configure live data stream to graphical output
  • 3JS dynamic data visualization schema.

Team Members and Division of Work

  • Matthew Draper: Stochastic Simulation Development
  • Timothy Beatham: Real-time data streaming from simulation to front end. Geolocation API configuration.
  • Billy Rooke: 3JS data & world map visualization.

Future Work

  • Run experiments & demonstrate framework for the program to run for an arbitrary town/locations public houses.
  • A object-oriented polymorphism approach to defining different agent archetypes and behaviours.
  • Modelling of group behaviours when agents enter probabilistic decision processes

Built With

Share this project:

Updates