This repository contains the scripts used in the practical component of my third year project. Stochastic Block Model with parameters number of cliques, clique size and re-wiring probability is used to generate a social network that has allows different opinions of agents to persist as well as intermix at the same time. the settings used for 100 clique are: clique size of 5 and a re-wiring probability of 0.05.
To perform your own simulations, initialise results directory to store simulation results. The only prerequisite is to have Python3 installed in your computer.
graph.py: Graph class to generate a Stochastic Block Model with the following parameters: number of cliques, clique size and re-wiring probability.simulation.py: script to run the main simulation on specified settings. Results stored in results/simulation.txt.structure.py: script to determine the structure (num. of terminals and non-terminals) of the networks generated across varying clique size and re-wiring probability for some fixed number of cliques. Results are stored seperately in results/structure_T.txt for terminals and in results/structure_nonT.txt for non-terminals.plot_simualtion.py: plots the initial active set against the resulting number of activations from the script simulation.py.plot_structure.py: plots the re-wiring probability + clique size against the number of terminals/non-terminals from the script structure.py. Need to specify which file to plot at the start.plot_model_features.py: plots the the number of terminals and non-terminals for a particular clique size against varying re-wiring probability. Linked with the results from the script structure.py.