To generate the graphs, run python Q1.py. The script saves the following graphs in the directory q1/px/, where x stands for the probability used in the simulation.

• Connectivity matrix of the excitory neurons (connectivity_matrix_x.svg)
• The firings scatter graph of excitory and inhibitory neuron layer (firings_x.svg)
• The mean firing rate of the 8 modules of excitory neurons (mean_firing_x.svg)

Plots for p = {0, 0.1, 0.2, 0.3, 0.4, 0.5} can be found in q1_example/.

To generate the graphs, run python Q2.py. The script saves the following graph in the directory q2/

• The multi-information plot for n number of trials (integration_n.svg)

Plots for n = {20, 50, 500} can be found in q2_example/.

Simulation of the modular networks was wrapped so parallelisation can be easily achieved using multiprocessing with the help of itertools. In both Q1.py and Q2.py, we have added try and except statements to proceed with the simulation with 1 core only should any exception occur (eg. failed to import multiprocessing module).

Izhikevich.py contains helper functions that is used in Q1 and Q2. GenerateNetwork creates an izhikevich neuron network. CompareMatrix compares 2 matrices and help in debugging differences in matrices. IzhikevichModularNetwork generates the connectivity matrix for the izhikevich network.Run.py contains the functions needed to perform the simulation on the given network. RunSimulation simulates the network for a given T and introduces background firing as well.