neuroidnet.ntr
Class Area

java.lang.Object
  |
  +--neuroidnet.ntr.Area

All Implemented Interfaces:

AreaInt, Expressive, Remote, Runnable, Serializable, Simulation

Direct Known Subclasses:

Area, ConceptArea, SensoryArea

public class Area

extends Object

implements Runnable, AreaInt, Serializable, Expressive, Simulation

Entity Area that holds neurons and keeps track of time.

Since:

1.0

Version:

1.0

Author:

Cengiz Gunay

See Also:

Serialized Form

Field Summary

(package private) double	activationThreshold The threshold to change a Neuroid from AM to AM1 mode.
protected Hashtable	axons Hash table holding the white matter, e.g.
(package private) int	conceptCount Count of concept allocations made (via call to addArbitrarySynapses) in this Area.
double	deltaT Increment of time for algorithms.
(package private) Synapse	fromInhibitorySynapseTemplate The template used for creating synapses from the inhibitoryInterNeuroid
int	id Id tag of the area.
(package private) boolean	inhibInter True if the area contains a inhibitory inter-neuroid.
(package private) Neuroid	inhibitoryInterNeuroid The global inhibitory neuroid that gets input from all and projects to all in the area.
(package private) AxonArbor	inhibitorySynapseVector The axon emanating from the inhibitoryInterNeuroid
(package private) Object	isCalculating Lock variable showing the thread is busy doing the calculations.
(package private) Object	isWaiting Lock variable showing the thread is waiting for a step request.
(package private) String	name Name of the Area for identification purposes.
Network	network Pointer to parent network TODO: Fix access modifier from public, problem with SensoryNeuroid
Vector	neuroids List of Neuroids contained in Area.
protected int	numberOfNeuroids Number of Neuroids contained in Area.
(package private) double	period Oscillation period of Neuroids.
(package private) int	replication The replication factor, i.e.
(package private) boolean	stepRequested Flag showing incoming requets to step().
(package private) Thread	thread Thread to do the actual calculations.
double	time Time of the area.
double	timeConstantM Membrane time constant of all neurons in this area.
(package private) Synapse	toInhibitorySynapseTemplate The template used for creating synapses to the inhibitoryInterNeuroid

Constructor Summary

Area(Network network, String name, int numberOfNeuroids, int replication, double period, double threshold, boolean inhibInter, double timeConstantM, double refractoryTimeConstant, Class neuroidClass)
Constructor with option to add inhibitory inter-neuron.

Area(Network network, String name, int numberOfNeuroids, int replication, double period, double threshold, double timeConstantM, double refractoryTimeConstant, Class neuroidClass)
Constructor for plain Area (no inhibitory interneuron).

Method Summary

void	addArbitrarySynapses(Synapse destSynapseTemplate, Neuroid srcNeuroid, int numberOfSynapses) Hack to arbitrarily choose different neurons for each allocation of concepts.
void	addAxon(Neuroid srcNeuroid, AxonArbor synapses) Adds the synapses to the outgoing synapse record of the Area.
int	addNeuroid(Neuroid neuroid) Adds a neuroid to the Area.
void	addRandomSynapses(Synapse destSynapseTemplate, Neuroid srcNeuroid, int numberOfSynapses) Creates new numberOfSynapses Synapses for a given srcNeuroid to a specified destination destArea.
void	connectToArea(AreaInt destArea, double timeConstantS, double delay, double nuBoost) Makes connections between this Area and the given destArea.
(package private) void	fireNeuroid(Neuroid neuroid) Fires neuroid.
double	getActivationThreshold() Get the value of activationThreshold.
Hashtable	getAxons() Get the value of axons.
String	getName() Get the value of name.
int	getNumberOfNeuroids() Get the value of numberOfNeuroids.
String	getProperties() Describe Area in more detail, including static properties.
Neuroid	getRandomNeuroid() Returns a Neuroid which is a random member of the destArea.
int	getReplication() Get the value of replication.
String	getStatus() Mentions the time in addition to toString() contents.
Thread	getThread() Get the value of thread.
void	init() Does nothing.
private void	readObject(ObjectInputStream in) Method called when a serialized object is loaded.
void	receiveSpike(Synapse synapse) Public method to receive spikes from other areas.
void	run() Check if step()ing requested, and serve the request.
void	setActivationThreshold(double v) Set the value of activationThreshold.
void	setAxons(Hashtable v) Set the value of axons.
void	setName(String v) Set the value of name.
void	setNumberOfNeuroids(int v) Set the value of numberOfNeuroids.
void	setReplication(int v) Set the value of replication.
void	setThread(Thread v) Set the value of thread.
void	step() Updates the state of the Area.
void	stop() Terminates the thread dedicated to this area.
String	toString() Identifies are with its name.
(package private) void	updateTime() Just increments the time by deltaT.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Field Detail

name

String name

Name of the Area for identification purposes.

time

public double time

Time of the area.

deltaT

public double deltaT

Increment of time for algorithms. Proxy for deltaT in network.

See Also:

Network.deltaT

id

public int id

Id tag of the area.

numberOfNeuroids

protected int numberOfNeuroids

Number of Neuroids contained in Area. TODO: should be obsolete, redundant info with size of neuroids vector.

See Also:

neuroids, Neuroid

replication

int replication

The replication factor, i.e. the number of Neuroids allocated for each concept.

period

double period

Oscillation period of Neuroids.

See Also:

Neuroid

neuroids

public Vector neuroids

List of Neuroids contained in Area. TODO: Fix access modifier from public, becuase of peripherals..

See Also:

Neuroid

inhibitoryInterNeuroid

Neuroid inhibitoryInterNeuroid

The global inhibitory neuroid that gets input from all and projects to all in the area.

fromInhibitorySynapseTemplate

Synapse fromInhibitorySynapseTemplate

The template used for creating synapses from the inhibitoryInterNeuroid

See Also:

inhibitoryInterNeuroid

toInhibitorySynapseTemplate

Synapse toInhibitorySynapseTemplate

The template used for creating synapses to the inhibitoryInterNeuroid

See Also:

inhibitoryInterNeuroid

inhibitorySynapseVector

AxonArbor inhibitorySynapseVector

The axon emanating from the inhibitoryInterNeuroid

See Also:

init()

axons

protected Hashtable axons

Hash table holding the white matter, e.g. the axons of neuroids residing in this Area connecting to Synapses of other Neuroids. Returns a Vector of Synapses for every Neuroid key.

See Also:

Synapse, Neuroid

isCalculating

transient volatile Object isCalculating

Lock variable showing the thread is busy doing the calculations.

See Also:

stepRequested, run(), step()

isWaiting

transient volatile Object isWaiting

Lock variable showing the thread is waiting for a step request.

See Also:

stepRequested, run(), step()

stepRequested

volatile boolean stepRequested

Flag showing incoming requets to step().

See Also:

run(), step()

thread

transient Thread thread

Thread to do the actual calculations. Is not part of the persistent object.

conceptCount

int conceptCount

Count of concept allocations made (via call to addArbitrarySynapses) in this Area.

See Also:

addArbitrarySynapses(neuroidnet.ntr.Synapse, neuroidnet.ntr.Neuroid, int), Peripheral.Concept

network

public Network network

Pointer to parent network TODO: Fix access modifier from public, problem with SensoryNeuroid

inhibInter

boolean inhibInter

True if the area contains a inhibitory inter-neuroid.

timeConstantM

public double timeConstantM

Membrane time constant of all neurons in this area.

activationThreshold

double activationThreshold

The threshold to change a Neuroid from AM to AM1 mode.

See Also:

Neuroid.step()

Constructor Detail

Area

public Area(Network network,
            String name,
            int numberOfNeuroids,
            int replication,
            double period,
            double threshold,
            double timeConstantM,
            double refractoryTimeConstant,
            Class neuroidClass)

Constructor for plain Area (no inhibitory interneuron). Calls other constructor.

Parameters:

network - a Network value

name - a String value

numberOfNeuroids - an int value

replication - an int value

period - a double value

threshold - a double value

See Also:

#Area(Network,String,int,int,double,double,boolean,double,double)

Area

public Area(Network network,
            String name,
            int numberOfNeuroids,
            int replication,
            double period,
            double threshold,
            boolean inhibInter,
            double timeConstantM,
            double refractoryTimeConstant,
            Class neuroidClass)

Constructor with option to add inhibitory inter-neuron. Creates a new Area instance with numberOfNeuroids of Neuroids.

TODO: If a allocate-on-demand approach is used for Neuroids none should be allocated at this time.

TODO: Automatically add areas to network?

Parameters:

network - a Network value

name - a String value

numberOfNeuroids - an int value

replication - an int value

period - a double value

threshold - a double value

inhibInter - a boolean value

See Also:

Neuroid

Method Detail

getName

public String getName()

Get the value of name.

Returns:

value of name.

setName

public void setName(String v)

Set the value of name.

Parameters:

v - Value to assign to name.

getNumberOfNeuroids

public int getNumberOfNeuroids()

Get the value of numberOfNeuroids.

Specified by:

getNumberOfNeuroids in interface AreaInt

Returns:

value of numberOfNeuroids.

setNumberOfNeuroids

public void setNumberOfNeuroids(int v)

Set the value of numberOfNeuroids.

Parameters:

v - Value to assign to numberOfNeuroids.

getReplication

public int getReplication()

Get the value of replication.

Specified by:

getReplication in interface AreaInt

Returns:

value of replication.

setReplication

public void setReplication(int v)

Set the value of replication.

Parameters:

v - Value to assign to replication.

getAxons

public Hashtable getAxons()

Get the value of axons.

Returns:

value of axons.

setAxons

public void setAxons(Hashtable v)

Set the value of axons.

Parameters:

v - Value to assign to axons.

getThread

public Thread getThread()

Get the value of thread.

Returns:

value of thread.

setThread

public void setThread(Thread v)

Set the value of thread.

Parameters:

v - Value to assign to thread.

getActivationThreshold

public double getActivationThreshold()

Get the value of activationThreshold.

Returns:

value of activationThreshold.

setActivationThreshold

public void setActivationThreshold(double v)

Set the value of activationThreshold.

Parameters:

v - Value to assign to activationThreshold.

addNeuroid

public int addNeuroid(Neuroid neuroid)

Adds a neuroid to the Area. Determines a sequence number according to the neuroid's order in the Vector neuroids. if specified in the constructor, a simple lateral circuit is formed. A link from this neuroid to the globally inhibitory neuroid is made.

Parameters:

neuroid - a Neuroid value to add into this Area

Returns:

the id of the neuroid (i.e. index in neuroids)

See Also:

Neuroid.id

connectToArea

public void connectToArea(AreaInt destArea,
                          double timeConstantS,
                          double delay,
                          double nuBoost)

Makes connections between this Area and the given destArea. Uses modified connection probability of random multipartite graphs to determine number of Neuroids to be connected on destination Area.

Specified by:

connectToArea in interface AreaInt

Parameters:

destArea - the Area to which this one is connected.

timeConstantS - time constant to be used in creating synapses

delay - synaptic delay

nuBoost - a double factor to multiply the original probability ratio as to magnify

See Also:

Neuroid, Synapse

addAxon

public void addAxon(Neuroid srcNeuroid,
                    AxonArbor synapses)

Adds the synapses to the outgoing synapse record of the Area. If synapses associated with srcNeuroid exist, new synapses are just added to them. TO DO: get a remote reference to an AxonArbor in the remote Area and put it in hash.

Parameters:

srcNeuroid - the presynaptic Neuroid in this Area.

synapses - the Vector to be associated with srcNeuroid

addRandomSynapses

public void addRandomSynapses(Synapse destSynapseTemplate,
                              Neuroid srcNeuroid,
                              int numberOfSynapses)

Creates new numberOfSynapses Synapses for a given srcNeuroid to a specified destination destArea. AxonArbor makes sure to return a set of synapses to distinct neurons (no repetitions!)

TODO: maybe put this method back into Area? addRandomSynapse should automatically add the synapse?

Specified by:

addRandomSynapses in interface AreaInt

Parameters:

destSynapseTemplate - a Synapse value

srcNeuroid - a Neuroid value

numberOfSynapses - an int value

addArbitrarySynapses

public void addArbitrarySynapses(Synapse destSynapseTemplate,
                                 Neuroid srcNeuroid,
                                 int numberOfSynapses)

Hack to arbitrarily choose different neurons for each allocation of concepts. A variable conceptCount is used to interleave the allocated neurons. Return a Vector of new numberOfSynapses Synapses. AxonArbor makes sure to return a set of synapses to distinct neurons (no repetitions!) Used from SensoryNeuroid.

Specified by:

addArbitrarySynapses in interface AreaInt

Parameters:

destSynapseTemplate - a Synapse value

srcNeuroid - a Neuroid value

numberOfSynapses - an int value

See Also:

conceptCount, periphery.SensoryNeuroid#SensoryNeuroid

getRandomNeuroid

public Neuroid getRandomNeuroid()

Returns a Neuroid which is a random member of the destArea.

TODO: one might might pseudo-random (organized) behavior in selecting destinations (like addArbitrarySynapses).

TODO: If a allocate-on-demand approach is used for neuroids, in case of finding a non-existing neuroid should result in its creation.

Returns:

a Neuroid value

See Also:

addArbitrarySynapses(neuroidnet.ntr.Synapse, neuroidnet.ntr.Neuroid, int), AxonArbor.destArea

receiveSpike

public void receiveSpike(Synapse synapse)

Public method to receive spikes from other areas. TO DO: get an AxonArbor and process synapses in the local area. OBSOLETE: redundant, instead use Synapse.receiveSpike() directly.

Parameters:

synapse - Synapse to spike.

See Also:

Synapse

updateTime

void updateTime()

Just increments the time by deltaT. Initial idea was to read current time from hardware clock. TODO: extend this class for the parallel version. (should we?)

init

public void init()

Does nothing. TODO: should do something.

Specified by:

init in interface Simulation

step

public void step()

Updates the state of the Area. Updates all Neuroids contained within. This method returns immediately, the thread does the calculation afterwards. TODO: Should be called from a thread with certain frequency. (should it?)

Specified by:

step in interface AreaInt

See Also:

Neuroid, run()

stop

public void stop()

Terminates the thread dedicated to this area.

Specified by:

stop in interface Simulation

See Also:

thread

fireNeuroid

void fireNeuroid(Neuroid neuroid)

Fires neuroid. Looks up synapses that its axon targets and calls receiveSpike method of all Synapses.

TODO: Make it trigger the Axon to fire.

Parameters:

neuroid - Neuroid to fire

See Also:

Neuroid, Synapse.receiveSpike()

toString

public String toString()

Identifies are with its name.

Specified by:

toString in interface Expressive

Overrides:

toString in class Object

Returns:

a String value

getStatus

public String getStatus()

Mentions the time in addition to toString() contents.

Specified by:

getStatus in interface Expressive

Returns:

a String value

See Also:

Expressive.toString()

getProperties

public String getProperties()

Describe Area in more detail, including static properties.

Specified by:

getProperties in interface Expressive

Returns:

a String value

See Also:

Expressive.getStatus()

run

public void run()

Check if step()ing requested, and serve the request.

Specified by:

run in interface Runnable

See Also:

step()

readObject

private void readObject(ObjectInputStream in)
                 throws IOException,
                        ClassNotFoundException

Method called when a serialized object is loaded. Only customization done is to create a new thread to reside in the transient thread variable.

Parameters:

in - a java.io.ObjectInputStream value

Throws:

IOException - if an error occurs

ClassNotFoundException - if an error occurs