Adding Projections from general BRAHMS processes to Populations of Neurons in a ModLIN System

A ModLIN system consists of 'populations' of rate coded neurons, connected via 'projections'. In order to build a system you must add populations, and then join them with projections.

To add a projection from a non-ModLIN process the nw_modAddInput() function is used. This function takes arguments that describe the projection that it is creating, and requires certain state data to be specified to the projection:

sys

This is the name of the SystemML system to add the projection to.

rate

The BRAHMS sample rate for the processes in the projection.

src

The BRAHMS process that is source of the projection.

src

The ModLIN destination population for the projection.

'path/to/proj'

The BRAHMS Namespace path of a ModLIN projection process.

projState

BRAHMS state data to be passed to the projection process. For the input this must include fields srcDims and srcLocs, which specify the dims and locs of the output of the source BRAHMS process.

setName

The set name to use to connect the projection process to the activation process of the destination population. This will determine how the input from the projection is intepreted by the population. For example, with the leaky activation process there are three types of synaptic connection, 'add' is additive (or subtractive if the weight is negative), 'shunt' is multiplicative, and 'shunt_div' is divisive.