Spike!
v1.0
A high speed Spiking Neural Network Simulator designed for GPGPUs.
|
#include <Synapses.h>
Public Attributes | |
int | max_number_of_connections_per_pair |
int | pairwise_connect_presynaptic |
int | pairwise_connect_postsynaptic |
int | gaussian_synapses_per_postsynaptic_neuron |
float | gaussian_synapses_standard_deviation |
float | weight_range_bottom |
float | weight_range_top |
float | random_connectivity_probability |
int | connectivity_type |
This struct accompanies the synapses class and is used to provide the a number of parameters to control synapse creation. This is also used as the foundation for the more sophisticated neuron parameter structs for SpikingSynapses etc.
A CONNECTIVITY_TYPE is set here
If using CONNECTIVITY_TYPE_GAUSSIAN_SAMPLE, should be set to the max number of desired connections to each postsynaptic neuron. Else, unused.
If using CONNECTIVITY_TYPE_GAUSSIAN_SAMPLE, should be set to the desired standard deviation of the gaussian distribution to be used to select presynaptic neurons. Else, unused.
Used in specific branches to indicate max number of connections between a specific pair of pre and postsynaptic neurons. UNUSED IN THIS BRANCH.
If using CONNECTIVITY_TYPE_ONE_TO_ONE, should be set to the postsynaptic neuron index. Else, unused.
If using CONNECTIVITY_TYPE_ONE_TO_ONE, should be set to the presynaptic neuron index. Else, unused.
If using CONNECTIVITY_TYPE_RANDOM, indicates the probability with which a connection should be made between the pre and post synaptic neurons. Else, unused.
Used for all CONNECTIVITY types. Indicates the lower range (inclusive) of the weight range to be used
Used for all CONNECTIVITY types. Indicates the upper range (inclusive) of the weight range to be used. If weight_range_bottom == weight_range_top, all weight values are equal. Else, they are a random uniform distribution in this range