cnslab.cnsnetwork

Class LayerStructure

java.lang.Object

cnslab.cnsnetwork.LayerStructure

public final class LayerStructure
extends Object

Organize layers of neurons.

Version:

$Date: 2012-08-04 13:43:22 -0500 (Sat, 04 Aug 2012) $ $Rev: 104 $ $Author: croft $

Author:

Yi Dong, David Wallace Croft

Field Summary

Fields

Modifier and Type	Field and Description
static int	ABMAP
Map<Integer,Axon>	axons
int	back_idum2
int[]	back_iv
int	back_iy
int	backseed
int	base
Map<String,TreeSet<Synapse>>[]	branchmaps Deprecated.
String[]	celltype Deprecated.
int	edgeLengthX
int	edgeLengthY
Seed	idum
Map<String,Integer>	indexmap for slow cellmap
List<Layer>	layerList
static int	LINEMAP
int	neuron_end
Neuron[]	neurons holds reference to neuron array after sent to network
int[]	nodeEndIndices end index in neuron for each node; .length=number parallel nodes
int	numberOfNeurons
int	numSYN
Map<String,Integer>	prefixSuffixToLayerIndexMap
Map<Integer,Map<String,SizeID>>	scaffold Scaffold is used to help to build up the network structure, which should be private
static int	sectionsInAxisX
static int	sectionsInAxisY

Constructor Summary

Constructors

Constructor and Description
LayerStructure(ModelFactory modelFactory, int nodeCount, int edgeLengthX, int edgeLengthY)

Method Summary

Methods

Modifier and Type	Method and Description
void	abmapcells_Main(int a, int b) Used in main Host, which is fast and used only to get numbers.
void	abmapcells(int a, int b, int host_id) Give the nethost id, map the cells in a times b, fast method.
void	addLayer(String pre, String suf, int xmult, int ymult, String type)
void	addLayer(String pre, String suf, int mult, String type) Add one layer of neurons into the structure.
void	buildStructure(int hostId) For nethost with id Hostid, build up the structure by using the scaffold build before.
int[]	celllayer_cordinate(int index, String pre, String suf) Deprecated.
boolean	celllayer_test(int index, String pre, String suf) Deprecated.
int	cellmap_slow(String pre, String suf, int x, int y) Deprecated.
String	celltype_slow(int index) Deprecated.
double	changeConnection(String f_pre, String f_suf, String t_pre, String t_suf, boolean convergent, double stre, int type, double delay, double[][] trans, int hostId, boolean periodic) Deprecated.
double	changeConnection(String f_pre, String f_suf, String t_pre, String t_suf, double stre, int type, double delay, double[][] trans, int hostId) Deprecated.
int[]	closePoint(int iterx, int itery, String tpre, String tsuf) Given tolayer of neuron and iterx, itery coordinates, return the closest point.
void	connect(String f_pre, String f_suf, String t_pre, String t_suf, boolean convergent, double stre, byte type, double delay, int idum) Deprecated.
double	connect(String f_pre, String f_suf, String t_pre, String t_suf, boolean convergent, double stre, int type, double delay, double[][] trans) Deprecated.
double	connect(String f_pre, String f_suf, String t_pre, String t_suf, boolean convergent, double stre, int type, double delay, double[][] proma, double[][] strma, int hostId, boolean periodic) Makeing connections between from and to layer of neurons, probability connection.
double	connect(String f_pre, String f_suf, String t_pre, String t_suf, boolean convergent, double stre, int type, double delay, double[][] trans, int hostId, boolean periodic) Makeing connections between from and to layer of neurons, L connection.
double	connect(String f_pre, String f_suf, String t_pre, String t_suf, double stre, int type, double delay, double[][] trans, int hostId) Makeing connections between from and to layer of neurons, general connection.
String	connectFrom(String id, PrintStream p) For debugging, pring out the connections connects from neuron with String id .
double	connectgp(String f_pre, String f_suf, String t_pre, String t_suf, double stre, int type, double delay, double[][] proma, double[][] strma, int hostId) Makeing connections between from and to layer of neurons, general probability connection.
Layer	getLayer(String prefix, String suffix)
int	getLayersize(String prefix) Find out the number of layers starting with prefix string
String	getLayerSuffix(String prefix, int index) Returns the suffix of the Nth layer with given prefix.
int	getXEdgeLength()
int	getXmultiplier(String prefix, String suffix)
int	getYEdgeLength()
int	getYmultiplier(String prefix, String suffix)
void	killMaps()
int	layerNumNeurons(String prefix, String suffix) Find out the number of neurons of a layer, given the layer name
void	mapcells() Map the cells in linear fashion.
void	neuronIndex(int hostId) set up the base neuron index for NetHost hostId
void	parseTarget(JpvmInfo info, SimulatorParser pas) Set up the target host (binary long type) information through network (JPVM).
void	popScaffold(String f_pre, String f_suf, String t_pre, String t_suf, boolean convergent, double stre, int type, double delay, double[][] proma, double[][] strma, int hostId, boolean periodic) Build up the Scaffold for later buiding up network structure.
void	popScaffold(String fromLayerPrefix, String fromLayerSuffix, String toLayerPrefix, String toLayerSuffix, boolean convergent, double strength, int type, double delay, double[][] trans, int hostId, boolean periodic) Build up the Scaffold for later building up network structure.
void	popScaffold(String f_pre, String f_suf, String t_pre, String t_suf, double stre, int type, double delay, double[][] proma, double[][] strma, int hostId) Build up the Scaffold for later buiding up network structure.
void	popScaffold(String fromLayerPrefix, String fromLayerSuffix, String toLayerPrefix, String toLayerSuffix, double strength, int type, double delay, double[][] trans, int hostId) Build up the Scaffold for later building up network structure.
void	qmapcells() Deprecated.
void	reSetSeed()
void	setSeed(Seed i)
void	sortSynapses(int hostId)
int	totalNumNeurons() Compute the total number of neurons

Methods inherited from class java.lang.Object

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

Field Detail

LINEMAP

public static final int LINEMAP

See Also:

Constant Field Values

ABMAP

public static final int ABMAP

See Also:

Constant Field Values

sectionsInAxisX

public static int sectionsInAxisX

sectionsInAxisY

public static int sectionsInAxisY

prefixSuffixToLayerIndexMap

public Map<String,Integer> prefixSuffixToLayerIndexMap

indexmap

public Map<String,Integer> indexmap

for slow cellmap

celltype

@Deprecated
public String[] celltype

Deprecated.

Remove maps after setup core!!!

branchmaps

@Deprecated
public Map<String,TreeSet<Synapse>>[] branchmaps

Deprecated.

scaffold

public Map<Integer,Map<String,SizeID>> scaffold

Scaffold is used to help to build up the network structure, which should be private

axons

public Map<Integer,Axon> axons

layerList

public List<Layer> layerList

edgeLengthX

public int edgeLengthX

edgeLengthY

public int edgeLengthY

numberOfNeurons

public int numberOfNeurons

nodeEndIndices

public int[] nodeEndIndices

end index in neuron for each node; .length=number parallel nodes

neurons

public Neuron[] neurons

holds reference to neuron array after sent to network

base

public int base

neuron_end

public int neuron_end

back_idum2

public int back_idum2

back_iy

public int back_iy

back_iv

public int[] back_iv

backseed

public int backseed

idum

public Seed idum

numSYN

public int numSYN

Constructor Detail

LayerStructure

public LayerStructure(ModelFactory modelFactory,
              int nodeCount,
              int edgeLengthX,
              int edgeLengthY)

Method Detail

neuronIndex

public void neuronIndex(int hostId)

set up the base neuron index for NetHost hostId

addLayer

public void addLayer(String pre,
            String suf,
            int mult,
            String type)

Add one layer of neurons into the structure.

Parameters:

pre -

suf -

mult -

type -

addLayer

public void addLayer(String pre,
            String suf,
            int xmult,
            int ymult,
            String type)

mapcells

public void mapcells()

Map the cells in linear fashion.

qmapcells

@Deprecated
public void qmapcells()

Deprecated.

abmapcells_Main

public void abmapcells_Main(int a,
                   int b)

Used in main Host, which is fast and used only to get numbers. neurons are not really initialized to save memory

Parameters:

a - sections in x axis

b - sections in y axis

cellmap_slow

@Deprecated
public int cellmap_slow(String pre,
                          String suf,
                          int x,
                          int y)

Deprecated.

Calculates the neuron index from the layer x,y coordinates.

celllayer_test

@Deprecated
public boolean celllayer_test(int index,
                                String pre,
                                String suf)

Deprecated.

celllayer_cordinate

@Deprecated
public int[] celllayer_cordinate(int index,
                                   String pre,
                                   String suf)

Deprecated.

celltype_slow

@Deprecated
public String celltype_slow(int index)

Deprecated.

Give the nethost id, map the cells in a times b, slow method.

Parameters:

a - parts in X dimension

b - parts in X dimension

host_id -

abmapcells

public void abmapcells(int a,
              int b,
              int host_id)

Give the nethost id, map the cells in a times b, fast method.

Parameters:

a - parts in X dimension

b - parts in X dimension

host_id -

totalNumNeurons

public int totalNumNeurons()

Compute the total number of neurons

Returns:

number of neurons

layerNumNeurons

public int layerNumNeurons(String prefix,
                  String suffix)

Find out the number of neurons of a layer, given the layer name

Parameters:

prefix - string

suffix - string

Returns:

number of neurons

getXmultiplier

public int getXmultiplier(String prefix,
                 String suffix)

getYmultiplier

public int getYmultiplier(String prefix,
                 String suffix)

getLayersize

public int getLayersize(String prefix)

Find out the number of layers starting with prefix string

Parameters:

prefix - string

Returns:

number of layers.

getLayer

public Layer getLayer(String prefix,
             String suffix)

getLayerSuffix

public String getLayerSuffix(String prefix,
                    int index)

Returns the suffix of the Nth layer with given prefix.

connect

@Deprecated
public void connect(String f_pre,
                      String f_suf,
                      String t_pre,
                      String t_suf,
                      boolean convergent,
                      double stre,
                      byte type,
                      double delay,
                      int idum)

Deprecated.

popScaffold

public void popScaffold(String fromLayerPrefix,
               String fromLayerSuffix,
               String toLayerPrefix,
               String toLayerSuffix,
               boolean convergent,
               double strength,
               int type,
               double delay,
               double[][] trans,
               int hostId,
               boolean periodic)

Build up the Scaffold for later building up network structure. L connection This method is called from class SimulatorParser method parseScaffoldWithPeriodic(int,Element).

popScaffold

public void popScaffold(String f_pre,
               String f_suf,
               String t_pre,
               String t_suf,
               boolean convergent,
               double stre,
               int type,
               double delay,
               double[][] proma,
               double[][] strma,
               int hostId,
               boolean periodic)

Build up the Scaffold for later buiding up network structure. L connection probability connection

Parameters:

f_pre -

f_suf -

t_pre -

t_suf -

convergent -

stre -

type -

delay -

trans -

hostId -

periodic -

popScaffold

public void popScaffold(String fromLayerPrefix,
               String fromLayerSuffix,
               String toLayerPrefix,
               String toLayerSuffix,
               double strength,
               int type,
               double delay,
               double[][] trans,
               int hostId)

Build up the Scaffold for later building up network structure. General connection

popScaffold

public void popScaffold(String f_pre,
               String f_suf,
               String t_pre,
               String t_suf,
               double stre,
               int type,
               double delay,
               double[][] proma,
               double[][] strma,
               int hostId)

Build up the Scaffold for later buiding up network structure. General population connection

Parameters:

f_pre -

f_suf -

t_pre -

t_suf -

convergent -

stre -

type -

delay -

trans -

hostId -

periodic -

buildStructure

public void buildStructure(int hostId)

For nethost with id Hostid, build up the structure by using the scaffold build before.

Parameters:

hostId -

sortSynapses

public void sortSynapses(int hostId)

connect

public double connect(String f_pre,
             String f_suf,
             String t_pre,
             String t_suf,
             boolean convergent,
             double stre,
             int type,
             double delay,
             double[][] trans,
             int hostId,
             boolean periodic)

Makeing connections between from and to layer of neurons, L connection.

Parameters:

f_pre -

f_suf -

t_pre -

t_suf -

convergent -

stre -

type -

delay -

trans -

hostId -

periodic -

Returns:

Largest weight

setSeed

public void setSeed(Seed i)

reSetSeed

public void reSetSeed()

connect

public double connect(String f_pre,
             String f_suf,
             String t_pre,
             String t_suf,
             boolean convergent,
             double stre,
             int type,
             double delay,
             double[][] proma,
             double[][] strma,
             int hostId,
             boolean periodic)

Makeing connections between from and to layer of neurons, probability connection.

Parameters:

f_pre -

f_suf -

t_pre -

t_suf -

convergent -

stre -

type -

delay -

trans -

hostId -

periodic -

Returns:

Largest weight

connect

public double connect(String f_pre,
             String f_suf,
             String t_pre,
             String t_suf,
             double stre,
             int type,
             double delay,
             double[][] trans,
             int hostId)

Makeing connections between from and to layer of neurons, general connection.

Parameters:

f_pre -

f_suf -

t_pre -

t_suf -

convergent -

stre -

type -

delay -

trans -

hostId -

periodic -

Returns:

Largest weight

connectgp

public double connectgp(String f_pre,
               String f_suf,
               String t_pre,
               String t_suf,
               double stre,
               int type,
               double delay,
               double[][] proma,
               double[][] strma,
               int hostId)

Makeing connections between from and to layer of neurons, general probability connection.

Parameters:

f_pre -

f_suf -

t_pre -

t_suf -

convergent -

stre -

type -

delay -

trans -

hostId -

periodic -

Returns:

Largest weight

changeConnection

@Deprecated
public double changeConnection(String f_pre,
                                 String f_suf,
                                 String t_pre,
                                 String t_suf,
                                 boolean convergent,
                                 double stre,
                                 int type,
                                 double delay,
                                 double[][] trans,
                                 int hostId,
                                 boolean periodic)

Deprecated.

changeConnection

@Deprecated
public double changeConnection(String f_pre,
                                 String f_suf,
                                 String t_pre,
                                 String t_suf,
                                 double stre,
                                 int type,
                                 double delay,
                                 double[][] trans,
                                 int hostId)

Deprecated.

connect

@Deprecated
public double connect(String f_pre,
                        String f_suf,
                        String t_pre,
                        String t_suf,
                        boolean convergent,
                        double stre,
                        int type,
                        double delay,
                        double[][] trans)

Deprecated.

closePoint

public int[] closePoint(int iterx,
               int itery,
               String tpre,
               String tsuf)

Given tolayer of neuron and iterx, itery coordinates, return the closest point. Called from popScaffold().

Parameters:

iterx -

itery -

tpre -

tsuf -

Returns:

connectFrom

public String connectFrom(String id,
                 PrintStream p)

For debugging, pring out the connections connects from neuron with String id . The format is Prefix,x,y,Suffix

Parameters:

Neuron - string

print - stream p

Returns:

Information.

killMaps

public void killMaps()

getXEdgeLength

public int getXEdgeLength()

getYEdgeLength

public int getYEdgeLength()

parseTarget

public void parseTarget(JpvmInfo info,
               SimulatorParser pas)
                 throws jpvmException

Set up the target host (binary long type) information through network (JPVM).

Parameters:

info -

pas -

Throws:

jpvmException