Data Structures

The LSNM simulator is contained in the python script sim.py. The data structure used to store the LSNM network, called modules, is a python dictionary containing the name of the module (a character string) and a description of the module (all parameters known about the specific module):

Key	Dictionary entry
Module name	Module description

Each entry in the modules dictionary depicted above, the module description, is a python list of lists containing number of columns and rows of the given module (matrix of units), activation rule (clamp or Wilson-Cowan), equation parameters needed to solve equations and find current values of module units, and the content of each unit in the module (itself a python list):

index 0	index 1	index 2	index 3-7	index 8
Number of columns	Number of rows	Activation rule	equation parameters (threshold, Delta, delta, K, N)	content of units

The content of units (index 8 in the table above) is a python list (usually a list of lists representing a matrix of nun elements). Each one of the elements of the content of units list contains another list with the following elements: the neural activity of the given unit at the current timestep, the sum of all inputs to the given unit, the sum of the inhibitory inputs to the given unit, the sum of the excitatory inputs to the given unit, and the outgoing connections (itself a python list) to other units in the network:

index 0	index 1	index 2	index 3	index 4
Neural activity	Sum of all inputs	Sum of excitatory inputs	Sum of inhibitory inputs	Outgoing connections

The outgoing connections (index 4 in the table above) for each given unit of a module is a python list containing 4-item lists with the following information: name of destination module (a string), column and row coordinates of destination unit within the destination module, and the value of the connection weight (see table below).

index 0	index 1	index 2	index 3
destination module	column of destination unit	row of destination unit	connection weight