SchellingsModelSimulation

A simulation of the Schelling's Segregation Model

Grid.py

Outlines the class Grid which is a representation of the simulation's grid.

attributes:

• size: the size of the grid, where the grid is of dimensions size x size.
• p: the fraction of neighbors each agent would like to be of the same type.
• q: the fraction of type 1 agents out of all agents.
• grid: an array representation of the grid, where 0 indicates an empty cell and 1 and 2 indicate agents of types 1 and 2 respectively.

methods:

init

Class constructor.

args:

• p: the fraction of neighbors each agent would like to be of the same type- should be in range [0, 1].
• q: the fraction of type 1 agents out of all agents- should also be in range [0, 1].
• size: the size of the grid, where the grid is of dimensions size x size. Default value: 50.
• empty: the fraction of empty cells out of all cells- should be in range [0, 1]. Default value: 0.1.

get_sim_pct

Calculates similarity percentage in neighborhood.

args:

• x: the row of the neighborhood center.
• y: the column of the neighborhood center.
• t: type to check against- if 0, then check against type of agent in the center. Default value: 0.

get_avg_sim

Calculates the average similarity percentage across neighborhoods in the entire grid. Return True if agent at location is unsatisfied, and False if agent is satisfied or no agent at location.

is_unsatisfied

Checks whether agent is unsatisfied.

args:

• x: the row of agent.
• y: the column of agent.

get_dist_cells

Gets coordinates of all cells in grid at the given Chebyshev distance.

args:

• x: the row of origin cell.
• y: the column of origin cell.
• distance: the distance

find_nearest_empty

Finds the nearest empty cell in which the agent would be satisfied.

args:

• x: the row of agent cell.
• y: the column of agent cell.

step

Takes a step in the simulation: chooses a random unsatisfied agent, places it in the nearest empty spot in which it'll be satisfied. Return True if it is able to complete the step, and False otherwise.

simulate

Runs the simulation.

args:

• max_steps: maximum number of steps to take before termination. If None, run until self.step() returns False.

simulation:

We ran the simulation for q = 0.5 and default values for size and empty for 100 equally spaced values of p in the range [0, 1].