Code for "Minimax Dynamics of Optimally Balanced Spiking Networks of Excitatory and Inhibitory Neurons". The four subfolders contains codes/data needed to reproduce each figure in the main text.
- Figure 1:
- To reproduce panels in figure 1 with data stored in Data folder
plot_figure_1_panels.m- To simulate tightly balanced network with a given
$\tau$
[rate,re,ri,W_EE,W_EI,W_IE,W_II,F,NE,N,tau] = balanced_network(tau)- For firing rate prediction accuracy
ropt = fr_prediction(tau,rate,re,ri,W_EE,W_EI,W_IE,W_II,F,NE,N) - Figure 2:
- To reproduce panels in figure 2 with data stored in Data folder
plot_figure_2_panels.m- To simulate network for signal reconstruction of spike trains with constant firing rate, with NE excitatory neurons
signal_reconstruction_synthetic(NE,0)- To simulate network for signal reconstruction of time varying input
signal_reconstruction_synthetic(NE,1)- To simulate network for signal reconstruction of natural image patches n with NI inhibitory neurons
signal_reconstruction_image(n,NI) - Figure 3:
- To reproduce panels in figure 3 with data stored in Data folder
plot_figure_3_panels.m- To simulate network with learned weights run
fixed_point_attractor(index, noise) - Figure 4:
- To reproduce panels in figure 4 with data stored in Data folder
plot_figure_4_panels.m- To simulate network that stores ring attractor
spiking_ring_attractor(theta_1,theta_2,gamma)- Simulate network for grid attractor
spiking_grid_attractor()
For meanings of the parameters in the above example code type
help function_name