---

Network Class Reference

This class manages the NN as it runs. More...

#include <amygdala/network.h>

Inheritance diagram for Network:

List of all members.

Public Types
typedef hash_map< AmIdInt, Neuron * >::const_iterator	const_iterator
typedef hash_map< unsigned int, Layer * >::iterator	layer_iterator
Public Methods
	Network ()
virtual	~Network ()
void	ScheduleNEvent (NEvent eventType, AmTimeInt eventTime, Neuron *reqNrn)
void	ScheduleNEvent (NEvent eventType, AmTimeInt eventTime, AmIdInt reqNrnId)
Neuron *	AddNeuron (NeuronRole nRole, AmIdInt nId)
Neuron *	AddNeuron (NeuronRole nRole, Neuron *nrn)
bool	ConnectNeurons (AmIdInt preSynapticNeuron, AmIdInt postSynapticNeuron, float weight, AmTimeInt delay=0)
bool	ConnectNeurons (Neuron *preSynapticNeuron, Neuron *postSynapticNeuron, float weight, AmTimeInt delay=0)
void	Run (AmTimeInt maxRunTime)
int	Size ()
int	LayerCount ()
const_iterator	begin () const
const_iterator	end () const
layer_iterator	layer_begin ()
layer_iterator	layer_end ()
void	AddLayer (Layer *newLayer)
AmIdInt	MaxNeuronId ()
bool	Layered ()
Neuron *	GetNeuron (AmIdInt nId)
SpikeInput *	GetSpikeInput ()
void	SetSpikeInput (SpikeInput *sIn)
bool	StreamingInput ()
void	StreamingInput (bool streaming)
void	SetTrainingMode (bool tMode)
bool	GetTrainingMode () const
Layer *	GetLayer (AmIdInt layerId)
void	EnableSpikeBatching ()
Static Public Methods
void	SetTimeStepSize (AmTimeInt stepSize)
AmTimeInt	TimeStepSize ()
AmTimeInt	SimTime ()
void	ResetSimTime ()
Protected Methods
void	ScheduleSpikeDelay (vector< Synapse * > &axon)
virtual void	IncrementSimTime ()
void	IncrementDelayOffset ()
Protected Attributes
int	pspLSize
int	pspLRes
int	netSize
bool	streamingInput
AmIdInt	nextNeuronId
unsigned int	nextLayerId
FunctionLookup *	functionRef
SpikeInput *	spikeInput
hash_map< AmIdInt, Neuron * >	net
hash_map< unsigned int, Layer * >	layers
priority_queue< SpikeRequest >	eventQ
priority_queue< SpikeRequest >	inputQ
vector< vector< Synapse * > >	delayedSpikeQ
AmTimeInt	currSpikeDelayOffset
AmTimeInt	maxOffset
bool	spikeDelaysOn
AmTimeInt	maxSpikeDelay
Synapse *	maxSpikeDelaySyn
unsigned int	spikeBatchCount
Static Protected Attributes
AmTimeInt	simStepSize
AmTimeInt	simTime
unsigned int	runCount
Friends
class	Neuron
void	MpSpikeInput::ReadInputBuffer ()

---

Detailed Description

This class manages the NN as it runs.

Network acts as a container for Neurons and Layers and also coordinates spike transmission between neurons. Amygdala supports multi-threading through the use of the MpNetwork class which allows neural nets to be partitioned across multiple Network objects. MpNetwork handles all communication between Networks and Neurons running in separate threads. This model will be extended in the future to allow Amygdala networks to run on clustered computer systems.

If there is a need to run two or more entirely separate neural nets at the same time, it is recommended that they be put into separate processes using fork() (or any other method that is useful). It would be possible to run separate networks by using MpNetwork, but the results may not be what is expected since MpNetwork is designed with the assumption that all MpNetwork objects belong to the same physical neural net.

See also:

MpNetwork

Author:

Matt Grover

---

Member Typedef Documentation

typedef hash_map<AmIdInt, Neuron*>::const_iterator Network::const_iterator

Define an iterator to parse through all Neurons contained in the network.

typedef hash_map<unsigned int, Layer*>::iterator Network::layer_iterator

Define an iterator to parse through all Layers contained in the network.

---

Member Function Documentation

void Network::AddLayer (  Layer *    newLayer )

Add a layer to the network. The Network assumes ownership over the Layer object and the Neuron objects contained within that Layer. Parameters: newLayer  Pointer to the Layer being added.

Neuron* Network::AddNeuron (  NeuronRole    nRole, Neuron *    nrn )

Add a Neuron to the network. Parameters: nRole  One of INPUTNEURON, HIDDENNEURON, or OUTPUTNEURON. nrn  Pointer to a Neuron. Network assumes ownership of the pointer once it has been added to the network. Returns: A pointer to the neuron is returned on success, a null pointer on failure.

Neuron* Network::AddNeuron (  NeuronRole    nRole, AmIdInt    nId )

Add a neuron to the network with a pre-determined neuron ID. Parameters: nRole  One of INPUTNEURON, HIDDENNEURON, or OUTPUTNEURON. nId  The desired neuron ID. Returns: A pointer to the new neuron on success. Null pointer on failure.

const_iterator Network::begin (    )  const [inline]

Returns: The first position of the iterator.

bool Network::ConnectNeurons (  Neuron *    preSynapticNeuron, Neuron *    postSynapticNeuron, float    weight, AmTimeInt    delay = 0 )

Connect two neurons together. Parameters: preSynapticNeuron  The pre-synaptic (originating) neuron pointer. postSynapticNeuron  The post-synaptic (receiving) neuron pointer. weight  A weight value in the range [-1, 1]. delay  The spike transmission delay in microseconds. Delay values will be rounded to the nearest whole multiple of the time step size. Returns: True on success.

bool Network::ConnectNeurons (  AmIdInt    preSynapticNeuron, AmIdInt    postSynapticNeuron, float    weight, AmTimeInt    delay = 0 )

Connect two neurons together. Parameters: preSynapticNeuron  The pre-synaptic (originating) neuron ID. postSynapticNeuron  The post-synaptic (receiving) neuron ID. weight  A weight value in the range [-1, 1]. delay  The spike transmission delay in microseconds. Delay values will be rounded to the nearest whole multiple of the time step size. Returns: True on success.

void Network::EnableSpikeBatching (    )  [inline]

Turn on spike batching (send all spikes to a neuron in a group, rather than one at a time). This is turned on automatically if spike delays are used and cannot be turned off again once this function has been called.

const_iterator Network::end (    )  const [inline]

Returns: The final position of the iterator.

Layer* Network::GetLayer (  AmIdInt    layerId )

Returns: a pointer to the Layer with layerID

Neuron* Network::GetNeuron (  AmIdInt    nId )  [inline]

Get a reference to a specific Neuron. Parameters: nId  The neuron ID of the Neuron. Returns: Pointer to the neuron with ID nId.

SpikeInput* Network::GetSpikeInput (    )  [inline]

Returns: Pointer to the SpikeInput object that is being used.

bool Network::GetTrainingMode (    )  const [inline]

Returns: True if training is enabled.

virtual void Network::IncrementSimTime (    )  [protected, virtual]

Increment the simTime variable to the next time step. In multithreaded mode, simTime is not incremented until all Nodes have called IncrementSimTime(). Reimplemented in MpNetwork.

layer_iterator Network::layer_begin (    )  [inline]

Returns: The first position of the layer iterator.

layer_iterator Network::layer_end (    )  [inline]

Returns: The final position of the layer iterator.

int Network::LayerCount (    )  [inline]

Returns: The number of layers in the network.

bool Network::Layered (    )  [inline]

Returns: True if the Network contains Layers.

AmIdInt Network::MaxNeuronId (    )  [inline]

Returns: The largest neuron ID currently in the network.

void Network::ResetSimTime (    )  [static]

Reset the simulation time to 0.

void Network::Run (  AmTimeInt    maxRunTime )

Run the simulation. The simulation will continue until either the event queues are completely empty or the network has run for maximum alloted time. If streaming input or MP mode is in use, the simulation will continue until maxRunTime has been reached even if the event queues are empty. Parameters: maxRunTime  Maximum time to run before returning in microseconds.

void Network::ScheduleNEvent (  NEvent    eventType, AmTimeInt    eventTime, AmIdInt    reqNrnId )

Schedule an event using the Neuron's ID instead of a pointer to the Neuron. Parameters: eventType  The action that the neuron should carry out. eventTime  Time that the event should be triggered. reqNrnId  ID of neuron that will execute the event.

void Network::ScheduleNEvent (  NEvent    eventType, AmTimeInt    eventTime, Neuron *    reqNrn )

Schedule an event. Only SPIKE, and INPUTSPIKE are currently used for the eventType. Parameters: eventType  The action that the neuron should carry out. eventTime  Time that the event should be triggered. reqNrn  Pointer to the neuron that will execute the event.

void Network::ScheduleSpikeDelay (  vector< Synapse * > &    axon )  [protected]

Schedule the transmission of a spike down an axon. This may be done in order to implement spike batching or to model transmission delays. This is normally called from Neuron. Parameters: axon  The axon vector from a Neuron. A spike will be scheduled to cross each Synapse after the delay time has passed. Delay times are stored in Synapse and set when Neurons are connected together. See also: Neuron::SendSpike(), Network::ConnectNeurons().

void Network::SetSpikeInput (  SpikeInput *    sIn )

Specify a SpikeInput object for the Network to use. Network creates a SimpleSpikeInput object in the constructor, so there is no need to call this function unless a different SpikeInput class is needed. Network assumes ownership of the pointers and the existing SpikeInput object is destroyed when a new one is passed in. Parameters: sIn  Pointer to the SpikeInput object.

void Network::SetTimeStepSize (  AmTimeInt    stepSize )  [static]

Set the size of the simulation time steps. This must be set before Neurons are added. Parameters: stepSize  Time step size in microseconds. The defaults to 100 us.

void Network::SetTrainingMode (  bool    tMode )

Toggle the training mode. Parameters: tMode  True if training should be enabled.

AmTimeInt Network::SimTime (    )  [inline, static]

Returns: The current simulation time in microseconds.

int Network::Size (    )  [inline]

Returns: Number of neurons in the network.

void Network::StreamingInput (  bool    streaming )  [inline]

Toggle the streaming input state. Parameters: streaming  A value of true turns on streaming input. Streaming input should be used if the input spike queue cannot be filled before calling Network::Run().

bool Network::StreamingInput (    )  [inline]

Returns: True if streaming input is in use.

AmTimeInt Network::TimeStepSize (    )  [inline, static]

Returns: The size of the time step in microseconds.

---

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

• network.h

---