Summary
PyGENESIS is a neural simulator with an interface to GENESIS 3. This project focuses on clarity and simplicity in use.
This simulator basically is an equation system solver with a transparent abstraction layer. Main idea in this abstraction is as below.
1. A component stores its own symbolic equation system.
2. Assembly of components conjoins the systems of each.
3. Simulation of a component is given out of the numerical solution of a system.
The evolution of an equation system can be closely monitored in predefined ways.
Motivation
Models implemented on popular neural simulators today are guaranteed power and efficiency that far exceeds that of custom implemented models. However, majority of neural simulations published today are custom implemented.
Neural simulators offer users ease in building models, through a tight concealment of mathematical internals. This good deal of abstraction however, acts as a hinderance to the inspection of internals.
Many researchers tend to implement customized models of their own for studying specific aspects of a model. Transparency of these models are gained at the expense of loss of scalability.
PyGENESIS aims to serve as a simulator for building transparent and scalable neuron models in a clear, pythonic way.
Goal
A concise, reliable, stand-alone neural simulator with an interface to GENESIS 3.
|