Mvaspike

Introduction

Mvaspike is a general purpose tool aimed at modeling and simulating large, complex networks of biological neural networks. It is based on an event-based modeling and simulation strategy, targetting mainly pulse-coupled, spiking neural networks (e.g. integrate-and-fire neurons, other spiking point neurons). We aimed at achieving a good balance between simulation efficiency and modeling freedom (complexity of models, extensibility, integration with other tools...).
Main features:

Networks:
- Delayed connections
- Populations of neurons
- Modular or hierarchical modeling strategy (borrowed from the DEVS formalism)
Neurons:
- Integrate-and-fire with STDP
- Stochastic neurons (work in progress)
- Phase-coded neurons
Simulation:
- Event-driven
Integration:
- Python bindings
- Standard file formats (e.g. XML and hdf5, work in progress)

Development

Everything is here.