GeomLib::AbstractNeuralDynamics Class Reference

#include <libs/GeomLib/AbstractNeuralDynamics.hpp>

Inheritance diagram for GeomLib::AbstractNeuralDynamics:

Collaboration diagram for GeomLib::AbstractNeuralDynamics:

Public Member Functions
	AbstractNeuralDynamics (const OdeParameter &, const CurrentCompensationParameter &=CurrentCompensationParameter(0., 0.))
virtual	~AbstractNeuralDynamics ()=0
virtual AbstractNeuralDynamics *	Clone () const =0
virtual Potential	EvolvePotential (Potential, MPILib::Time) const =0
virtual std::vector< Potential >	InterpretationArray () const =0
const OdeParameter &	Par () const
const CurrentCompensationParameter &	ParCur () const
virtual MPILib::Time	TPeriod () const =0
virtual MPILib::Time	TStep () const =0

Protected Attributes
const OdeParameter	_par
const CurrentCompensationParameter	_par_current

Detailed Description

The configuration of a GeomAlgorithm requires that the neural dynamics is defined somewhere. The dynamics is used to define an OdeSystem.

There are predefined derived classes for leaky-integrate-and-fire and quadratic-integrate-and-fire dynamics. The dynamics is usually defined on a grid, whose dimensions are specified in the OdeParameter. OdeParameter also contains a NeuronParameter that determines the dynamics.Anyone who wants to use their own model of neural dynamics. To define dynamics on this grid, the EvolvePotential must be overloaded. Derived classes exist for LIF and QIF neurons, and more generally for spiking neurons (SpikingNeuralDynamics).

Definition at line 39 of file AbstractNeuralDynamics.hpp.

Constructor & Destructor Documentation

AbstractNeuralDynamics::AbstractNeuralDynamics	(	const OdeParameter &	par_ode,
		const CurrentCompensationParameter &	par_current = CurrentCompensationParameter(0.,0.)
	)

Constructor.

Parameters

par_current

Neuron parameter values, and other information relevant to binning structure By default, no current compensation, but if it's chosen, it directly affects the neural dynamics and hence the binning

Definition at line 26 of file AbstractNeuralDynamics.cpp.

AbstractNeuralDynamics::~AbstractNeuralDynamics ( )

pure virtual

Virtual destructor.

Definition at line 36 of file AbstractNeuralDynamics.cpp.

Member Function Documentation

virtual AbstractNeuralDynamics* GeomLib::AbstractNeuralDynamics::Clone ( ) const

pure virtual

virtual construction mechanism

Implemented in GeomLib::SpikingQifNeuralDynamics, GeomLib::LifNeuralDynamics, GeomLib::SpikingLifNeuralDynamics, and GeomLib::SpikingNeuralDynamics.

virtual Potential GeomLib::AbstractNeuralDynamics::EvolvePotential ( Potential  , MPILib::Time    ) const

pure virtual

Given a potential, specify how it evolves over a given time step The range of validity of this function is determined by the overloaded function.

virtual std::vector<Potential> GeomLib::AbstractNeuralDynamics::InterpretationArray ( ) const

pure virtual

Generate the bin boundaries for geometric binning based on the dyn.

Implemented in GeomLib::LifNeuralDynamics, and GeomLib::SpikingNeuralDynamics.

const OdeParameter& GeomLib::AbstractNeuralDynamics::Par ( ) const

inline

Provide efficient access. For use in time critical code.

Definition at line 65 of file AbstractNeuralDynamics.hpp.

References _par.

const CurrentCompensationParameter& GeomLib::AbstractNeuralDynamics::ParCur ( ) const

inline

Return the current compensation object; can be used to test whether current compensation is applied.

Definition at line 78 of file AbstractNeuralDynamics.hpp.

References _par_current.

virtual MPILib::Time GeomLib::AbstractNeuralDynamics::TPeriod ( ) const

pure virtual

Period for the dynamic model. Consult the '1D document'.

Implemented in GeomLib::SpikingQifNeuralDynamics, GeomLib::LifNeuralDynamics, and GeomLib::SpikingLifNeuralDynamics.

virtual MPILib::Time GeomLib::AbstractNeuralDynamics::TStep ( ) const

pure virtual

Fundamental time step by which mass is shifted through the geometric bins. Consult the '1D document' for details.

Implemented in GeomLib::SpikingQifNeuralDynamics, GeomLib::LifNeuralDynamics, and GeomLib::SpikingLifNeuralDynamics.

Member Data Documentation

const OdeParameter GeomLib::AbstractNeuralDynamics::_par

protected

Time critical access for derived classes.

Definition at line 83 of file AbstractNeuralDynamics.hpp.

Referenced by GeomLib::SpikingLifNeuralDynamics::EvolvePotential(), GeomLib::SpikingQifNeuralDynamics::EvolvePotential(), GeomLib::LifNeuralDynamics::InterpretationArray(), GeomLib::LifNeuralDynamics::Nneginit(), GeomLib::LifNeuralDynamics::Nposinit(), Par(), GeomLib::LifNeuralDynamics::TimePeriod(), GeomLib::SpikingLifNeuralDynamics::TimeToInf(), GeomLib::SpikingQifNeuralDynamics::TimeToInf(), GeomLib::SpikingLifNeuralDynamics::TPeriod(), GeomLib::SpikingQifNeuralDynamics::TPeriod(), GeomLib::SpikingLifNeuralDynamics::TStep(), and GeomLib::SpikingQifNeuralDynamics::TStep().

const CurrentCompensationParameter GeomLib::AbstractNeuralDynamics::_par_current

protected

For use in the concrete dynamics instantiation.

Definition at line 86 of file AbstractNeuralDynamics.hpp.

Referenced by ParCur().

---

The documentation for this class was generated from the following files:

• /Users/scsmdk/miind-git/libs/GeomLib/AbstractNeuralDynamics.hpp
• /Users/scsmdk/miind-git/libs/GeomLib/AbstractNeuralDynamics.cpp