SparseMatrixLinker Class Reference

SparseMatrixLinker Class extend MatrixLinker for allow non-full connection between Clusters. More...

Inheritance diagram for SparseMatrixLinker:

[legend]Collaboration diagram for SparseMatrixLinker:

[legend]List of all members.

Constructors
	SparseMatrixLinker (Cluster *from, Cluster *to, const char *name="unnamed")
	SparseMatrixLinker (Real prob, Cluster *from, Cluster *to, const char *name="unnamed")
	SparseMatrixLinker (Cluster *from, Cluster *to, Real prob, bool zeroDiagonal=false, bool symmetricMask=false, const char *name="unnamed")
	SparseMatrixLinker (PropertySettings &prop)
virtual	~SparseMatrixLinker ()
Interface
virtual void	setWeight (u_int from, u_int to, Real weight)
virtual void	randomize (Real min, Real max)
void	update ()
void	connect (u_int from, u_int to)
void	connectRandom (Real prob)
void	connectAll ()
void	disconnect (u_int from, u_int to)
void	disconnectRandom (Real prob)
void	disconnectAll ()
MatrixData< bool > &	getMask ()
MatrixData< bool > &	mask ()
Variant	maskP ()
void	setMask (const MatrixData< bool > &mask)
bool	setMask (const Variant &v)
virtual SparseMatrixLinker *	clone () const

---

Detailed Description

Every connection is weighted, and the weight is memorized into a weight-matrix Details ...

---

Constructor & Destructor Documentation

SparseMatrixLinker	(	Cluster *	from,
		Cluster *	to,
		const char *	name = "unnamed"
	)

Connect clusters with complete connections

SparseMatrixLinker	(	Real	prob,
		Cluster *	from,
		Cluster *	to,
		const char *	name = "unnamed"
	)

Connect neurons of Clusters with a random connections with the passed probability.

SparseMatrixLinker	(	Cluster *	from,
		Cluster *	to,
		Real	prob,
		bool	zeroDiagonal = false,
		bool	symmetricMask = false,
		const char *	name = "unnamed"
	)

Connect neurons of Clusters with a random connections with the passed probability.
With this contructor you must also specify whether the diagonal of the matrix is made of zeros and whether the matrix is symmetrical

Warning:

You can use this constructor only with square matrices, otherwise it will generate a memory error!!!

SparseMatrixLinker

(

PropertySettings &

prop

)

Construct by PropertySettings

virtual ~SparseMatrixLinker

(

)

[virtual]

Destructor

---

Member Function Documentation

virtual void setWeight	(	u_int	from,
		u_int	to,
		Real	weight
	)			[virtual]

Set the weight of the connection specified

Reimplemented from MatrixLinker.

virtual void randomize	(	Real	min,
		Real	max
	)			[virtual]

Randomize the weights of the SparseMatrixLinker

Reimplemented from MatrixLinker.

void update

(

)

[virtual]

Performs the dot-product calculation where the non-connection are considered as zero

Implements Updatable.

void connect	(	u_int	from,
		u_int	to
	)

Connect two neurons

void connectRandom

(

Real

prob

)

Connects randomly according to the given probability

void connectAll

(

)

Connect all couples of neurons

void disconnect	(	u_int	from,
		u_int	to
	)

Disconnect the two neurons

void disconnectRandom

(

Real

prob

)

Disconnects randomly according to the given probability

void disconnectAll

(

)

Disonnect all couples of neurons

MatrixData<bool>& getMask

(

)

[inline]

Get the mask

Deprecated:

use mask() instead

MatrixData<bool>& mask

(

)

[inline]

Return the mask

Variant maskP

(

)

[inline]

Return the mask matrix (Variant ver)

void setMask

(

const MatrixData< bool > &

mask

)

Set the whole mask matrix

bool setMask

(

const Variant &

v

)

Set the whole mask matrix (Variant ver)

virtual SparseMatrixLinker* clone

(

)

const [virtual]

Clone this SparseMatrixLinker

Reimplemented from Linker.

---

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

• include/sparsematrixlinker.h

Copyright © 2005-2008 G. Massera

Contact: emmegian (at) yahoo (dot) it

  Documentation generated by