graph LR
bindsnet_network_network["bindsnet.network.network"]
bindsnet_network_nodes["bindsnet.network.nodes"]
bindsnet_network_topology["bindsnet.network.topology"]
bindsnet_network_network -- "composes and manages" --> bindsnet_network_nodes
bindsnet_network_network -- "drives the simulation of" --> bindsnet_network_nodes
bindsnet_network_network -- "composes and manages" --> bindsnet_network_topology
bindsnet_network_network -- "drives the simulation of" --> bindsnet_network_topology
bindsnet_network_nodes -- "generates spikes for" --> bindsnet_network_topology
bindsnet_network_topology -- "delivers inputs to" --> bindsnet_network_nodes
The bindsnet.network subsystem forms the core of the Spiking Neural Network simulation. The bindsnet.network.network component acts as the central orchestrator, responsible for the overall execution and training of the SNN. It integrates and manages instances of bindsnet.network.nodes, which define the behavior of individual neurons, and bindsnet.network.topology, which establishes and maintains the synaptic connections between these neurons. Spikes generated by bindsnet.network.nodes are propagated through the connections defined by bindsnet.network.topology, which in turn delivers inputs to other bindsnet.network.nodes, completing the neural network's operational cycle. This modular design allows for flexible construction and simulation of various SNN architectures.
This is the central control unit for the entire Spiking Neural Network. It is responsible for initializing the network, managing its execution flow (e.g., run method), and overseeing the training process (train method). It acts as the primary interface for interacting with and controlling the SNN.
Related Classes/Methods:
Defines the fundamental properties and dynamic behaviors of individual spiking neurons. It handles how neurons process inputs, manage their internal states (e.g., membrane potential, thresholds), and decay over time. This component supports various computational kernel types (e.g., EtaKernel, AlphaKernel) that dictate the neuron's response to spikes.
Related Classes/Methods:
Manages the structural organization and synaptic connections between different neuron populations within the network. It defines how neurons are interconnected, including the properties of synapses (e.g., weights, delays) and how spikes propagate through these connections.
Related Classes/Methods: