PLearn::ReconstructionWeightsKernel Class Reference

#include <ReconstructionWeightsKernel.h>

Inheritance diagram for PLearn::ReconstructionWeightsKernel:

[legend]Collaboration diagram for PLearn::ReconstructionWeightsKernel:

[legend]List of all members.


Public Member Functions
	ReconstructionWeightsKernel ()
	Default constructor.
virtual void	build ()
	Simply calls inherited::build() then build_().
virtual void	makeDeepCopyFromShallowCopy (map< const void , void > &copies)
	Transforms a shallow copy into a deep copy.
	PLEARN_DECLARE_OBJECT (ReconstructionWeightsKernel)
virtual real	evaluate (const Vec &x1, const Vec &x2) const
	Compute K(x1,x2).
virtual real	evaluate_i_j (int i, int j) const
	returns evaluate(data(i),data(j))
virtual real	evaluate_i_x (int i, const Vec &x, real squared_norm_of_x=-1) const
virtual real	evaluate_x_i (const Vec &x, int i, real squared_norm_of_x=-1) const
	returns evaluate(x,data(i)) [default version calls evaluate_i_x if kernel is_symmetric]
virtual real	evaluate_x_i_again (const Vec &x, int i, real squared_norm_of_x=-1, bool first_time=false) const
virtual void	setDataForKernelMatrix (VMat the_data)
	Subclasses may overload these methods to provide efficient kernel matrix access
virtual real	evaluate_sum_k_i_k_j (int i, int j) const
	Return sum_k K(x_k, x_i) * K(x_k, x_j).
virtual void	computeLLEMatrix (const Mat &lle_mat) const
	Fill 'lle_mat', which must be of size (n x n), with entries (i,j) equal to W_{ij} + W_{ji} - W_{ki} W_{kj} (this is used in LLE to compute the kernel Gram matrix).
Public Attributes
Ker	distance_kernel
Ker	dot_product_kernel
int	ignore_nearest
int	knn
real	regularizer
Protected Member Functions
void	computeWeights ()
	Precompute the weights W_{ij}.
void	reconstruct (const Vec &x, const TVec< int > &neighbors, Vec &w) const
	Compute the reconstruction weights for a vector, given its nearest neighbors.
Static Protected Member Functions
void	declareOptions (OptionList &ol)
	Declares this class' options.
Protected Attributes
Ker	dist_ker
	The kernel used to compute distances (equal to 'distance_kernel' if specified, and otherwise the usual Euclidean distance).
Ker	dp_ker
	The kernel used to compute dot products (equal to 'dot_product_kernel' if specified, and otherwise the usual DotProductKernel).
TMat< int >	neighbors
	The indices of the neighbors of each data point.
TVec< TMat< int > >	is_neighbor_of
	The element i is a matrix whose first column is the list of the points which have i among their neighbors, and the second column is the index of i in these neighbors.
PP< SelectRowsVMatrix >	sub_data
	Points toward a subset of the training data (typically, a neighborhood).
Mat	weights
	The matrix with the weights W_{ij}.
Private Types
typedef Kernel	inherited
Private Member Functions
void	build_ ()
	This does the actual building.
Private Attributes
bool	build_in_progress
	True iff build() has been called but build_() has not been called yet.
bool	new_data
	True iff the 'setDataForKernelMatrix' method has been called since last time we called 'reconstruct'.
Mat	k_xi_x_sorted
	Used in 'evaluate_x_i_again' to store the distances from x to its nearest neighbors.
TVec< int >	neighbors_of_x
	Used in 'evaluate_x_i_again' to store the neighbors of x.
Vec	weights_x
	Used in 'evaluate_x_i_again' to store the reconstruction weights of x.
Mat	local_gram
	Used in 'reconstruct' to store the local Gram matrix.
PP< ShiftAndRescaleVMatrix >	centered_neighborhood
	Used in 'reconstruct' to point toward the locally centered data.
Vec	ones
	Used in 'reconstruct' to store a vector filled with 1.

Member Typedef Documentation

typedef Kernel PLearn::ReconstructionWeightsKernel::inherited [private]

Reimplemented from PLearn::Kernel.
Definition at line 59 of file ReconstructionWeightsKernel.h.

Constructor & Destructor Documentation

PLearn::ReconstructionWeightsKernel::ReconstructionWeightsKernel ( )

Default constructor.

Definition at line 55 of file ReconstructionWeightsKernel.cc.
References sub_data.

Member Function Documentation

void PLearn::ReconstructionWeightsKernel::build ( ) [virtual]

Simply calls inherited::build() then build_().

Reimplemented from PLearn::Kernel.
Definition at line 112 of file ReconstructionWeightsKernel.cc.
References build_(), and build_in_progress.

void PLearn::ReconstructionWeightsKernel::build_ ( ) [private]

This does the actual building.

Reimplemented from PLearn::Kernel.
Definition at line 122 of file ReconstructionWeightsKernel.cc.
References build_in_progress, dist_ker, distance_kernel, dot_product_kernel, dp_ker, ignore_nearest, knn, PLERROR, and setDataForKernelMatrix().
Referenced by build().

void PLearn::ReconstructionWeightsKernel::computeLLEMatrix ( const Mat & lle_mat ) const [virtual]

Fill 'lle_mat', which must be of size (n x n), with entries (i,j) equal to W_{ij} + W_{ji} - W_{ki} W_{kj} (this is used in LLE to compute the kernel Gram matrix).

Definition at line 151 of file ReconstructionWeightsKernel.cc.
References PLearn::TMat< T >::clear(), k, knn, PLearn::TMat< T >::length(), PLearn::Mat, neighbors, PLERROR, PLearn::ProgressBar::update(), weights, and PLearn::TMat< T >::width().

void PLearn::ReconstructionWeightsKernel::computeWeights ( ) [protected]

Precompute the weights W_{ij}.

Definition at line 181 of file ReconstructionWeightsKernel.cc.
References dist_ker, PLearn::VMat::getSubRow(), ignore_nearest, is_neighbor_of, knn, PLearn::Mat, neighbors, PLERROR, reconstruct(), PLearn::TVec< TMat< int > >::resize(), PLearn::TMat< T >::resize(), PLearn::TVec< T >::resize(), PLearn::sortRows(), PLearn::ProgressBar::update(), PLearn::Vec, and weights.
Referenced by setDataForKernelMatrix().

void PLearn::ReconstructionWeightsKernel::declareOptions ( OptionList & ol ) [static, protected]

Declares this class' options.

Reimplemented from PLearn::Kernel.
Definition at line 83 of file ReconstructionWeightsKernel.cc.
References PLearn::declareOption(), and PLearn::OptionList.

real PLearn::ReconstructionWeightsKernel::evaluate ( const Vec & x1,

const Vec & x2

) const [virtual]

Compute K(x1,x2).

Implements PLearn::Kernel.
Definition at line 229 of file ReconstructionWeightsKernel.cc.
References evaluate_x_i().

real PLearn::ReconstructionWeightsKernel::evaluate_i_j ( int i,

int j

) const [virtual]

returns evaluate(data(i),data(j))

Reimplemented from PLearn::Kernel.
Definition at line 243 of file ReconstructionWeightsKernel.cc.
References ignore_nearest, k, knn, neighbors, PLERROR, weights, and PLearn::TMat< T >::width().
Referenced by evaluate_i_x().

real PLearn::ReconstructionWeightsKernel::evaluate_i_x ( int i,

const Vec & x,

real squared_norm_of_x = -1

) const [virtual]

returns evaluate(data(i),x) [squared_norm_of_x is just a hint that may allow to speed up computation if it is already known, but it's optional]
Reimplemented from PLearn::Kernel.
Definition at line 270 of file ReconstructionWeightsKernel.cc.
References evaluate_i_j(), and x.

real PLearn::ReconstructionWeightsKernel::evaluate_sum_k_i_k_j ( int i,

int j

) const [virtual]

Return sum_k K(x_k, x_i) * K(x_k, x_j).

Definition at line 281 of file ReconstructionWeightsKernel.cc.
References ignore_nearest, is_neighbor_of, knn, PLearn::TMat< T >::length(), PLERROR, PLearn::sum(), weights, and PLearn::TMat< T >::width().

real PLearn::ReconstructionWeightsKernel::evaluate_x_i ( const Vec & x,

int i,

real squared_norm_of_x = -1

) const [virtual]

returns evaluate(x,data(i)) [default version calls evaluate_i_x if kernel is_symmetric]

Reimplemented from PLearn::Kernel.
Definition at line 319 of file ReconstructionWeightsKernel.cc.
References evaluate_x_i_again(), and x.
Referenced by evaluate().

real PLearn::ReconstructionWeightsKernel::evaluate_x_i_again ( const Vec & x,

int i,

real squared_norm_of_x = -1,

bool first_time = false

) const [virtual]

Reimplemented from PLearn::Kernel.
Definition at line 326 of file ReconstructionWeightsKernel.cc.
References dist_ker, PLearn::TVec< int >::find(), ignore_nearest, k_xi_x_sorted, knn, neighbors_of_x, reconstruct(), PLearn::TVec< int >::resize(), PLearn::TMat< T >::subMat(), weights_x, and x.
Referenced by evaluate_x_i().

void PLearn::ReconstructionWeightsKernel::makeDeepCopyFromShallowCopy ( map< const void *, void * > & copies ) [virtual]

Transforms a shallow copy into a deep copy.

Definition at line 345 of file ReconstructionWeightsKernel.cc.
References PLERROR.

PLearn::ReconstructionWeightsKernel::PLEARN_DECLARE_OBJECT ( ReconstructionWeightsKernel )

void PLearn::ReconstructionWeightsKernel::reconstruct ( const Vec & x,

const TVec< int > & neighbors,

Vec & w

) const [protected]

Compute the reconstruction weights for a vector, given its nearest neighbors.

Definition at line 362 of file ReconstructionWeightsKernel.cc.
References centered_neighborhood, dp_ker, PLearn::TVec< T >::fill(), PLearn::TVec< T >::length(), local_gram, new_data, ones, PLearn::regularizeMatrix(), regularizer, PLearn::TMat< T >::resize(), PLearn::TVec< T >::resize(), PLearn::solveLinearSystem(), sub_data, PLearn::sum(), weights_x, and x.
Referenced by computeWeights(), and evaluate_x_i_again().

void PLearn::ReconstructionWeightsKernel::setDataForKernelMatrix ( VMat the_data ) [virtual]

** Subclasses may overload these methods to provide efficient kernel matrix access **
This method sets the data VMat that will be used to define the kernel matrix. It may precompute values from this that may later accelerate the evaluation of a kernel matrix element
Reimplemented from PLearn::Kernel.
Definition at line 405 of file ReconstructionWeightsKernel.cc.
References build_in_progress, computeWeights(), dist_ker, new_data, and sub_data.
Referenced by build_().

Member Data Documentation

bool PLearn::ReconstructionWeightsKernel::build_in_progress [private]

True iff build() has been called but build_() has not been called yet.

Definition at line 62 of file ReconstructionWeightsKernel.h.
Referenced by build(), build_(), and setDataForKernelMatrix().

PP<ShiftAndRescaleVMatrix> PLearn::ReconstructionWeightsKernel::centered_neighborhood [mutable, private]

Used in 'reconstruct' to point toward the locally centered data.

Definition at line 83 of file ReconstructionWeightsKernel.h.
Referenced by reconstruct().

Ker PLearn::ReconstructionWeightsKernel::dist_ker [protected]

The kernel used to compute distances (equal to 'distance_kernel' if specified, and otherwise the usual Euclidean distance).

Definition at line 98 of file ReconstructionWeightsKernel.h.
Referenced by build_(), computeWeights(), evaluate_x_i_again(), and setDataForKernelMatrix().

Ker PLearn::ReconstructionWeightsKernel::distance_kernel

Definition at line 125 of file ReconstructionWeightsKernel.h.
Referenced by build_().

Ker PLearn::ReconstructionWeightsKernel::dot_product_kernel

Definition at line 126 of file ReconstructionWeightsKernel.h.
Referenced by build_().

Ker PLearn::ReconstructionWeightsKernel::dp_ker [protected]

The kernel used to compute dot products (equal to 'dot_product_kernel' if specified, and otherwise the usual DotProductKernel).

Definition at line 102 of file ReconstructionWeightsKernel.h.
Referenced by build_(), and reconstruct().

int PLearn::ReconstructionWeightsKernel::ignore_nearest

Definition at line 127 of file ReconstructionWeightsKernel.h.
Referenced by build_(), computeWeights(), evaluate_i_j(), evaluate_sum_k_i_k_j(), and evaluate_x_i_again().

TVec< TMat<int> > PLearn::ReconstructionWeightsKernel::is_neighbor_of [protected]

The element i is a matrix whose first column is the list of the points which have i among their neighbors, and the second column is the index of i in these neighbors.
The first column is sorted by increasing index, and does not contain i.
Definition at line 111 of file ReconstructionWeightsKernel.h.
Referenced by computeWeights(), and evaluate_sum_k_i_k_j().

Mat PLearn::ReconstructionWeightsKernel::k_xi_x_sorted [mutable, private]

Used in 'evaluate_x_i_again' to store the distances from x to its nearest neighbors.

Definition at line 71 of file ReconstructionWeightsKernel.h.
Referenced by evaluate_x_i_again().

int PLearn::ReconstructionWeightsKernel::knn

Definition at line 128 of file ReconstructionWeightsKernel.h.
Referenced by build_(), computeLLEMatrix(), computeWeights(), evaluate_i_j(), evaluate_sum_k_i_k_j(), and evaluate_x_i_again().

Mat PLearn::ReconstructionWeightsKernel::local_gram [mutable, private]

Used in 'reconstruct' to store the local Gram matrix.

Definition at line 80 of file ReconstructionWeightsKernel.h.
Referenced by reconstruct().

TMat<int> PLearn::ReconstructionWeightsKernel::neighbors [protected]

The indices of the neighbors of each data point.

Definition at line 105 of file ReconstructionWeightsKernel.h.
Referenced by computeLLEMatrix(), computeWeights(), and evaluate_i_j().

TVec<int> PLearn::ReconstructionWeightsKernel::neighbors_of_x [mutable, private]

Used in 'evaluate_x_i_again' to store the neighbors of x.

Definition at line 74 of file ReconstructionWeightsKernel.h.
Referenced by evaluate_x_i_again().

bool PLearn::ReconstructionWeightsKernel::new_data [mutable, private]

True iff the 'setDataForKernelMatrix' method has been called since last time we called 'reconstruct'.
This is necessary to ensure everything is correctly initialized in the 'reconstruct' method.
Definition at line 67 of file ReconstructionWeightsKernel.h.
Referenced by reconstruct(), and setDataForKernelMatrix().

Vec PLearn::ReconstructionWeightsKernel::ones [mutable, private]

Used in 'reconstruct' to store a vector filled with 1.

Definition at line 86 of file ReconstructionWeightsKernel.h.
Referenced by reconstruct().

real PLearn::ReconstructionWeightsKernel::regularizer

Definition at line 129 of file ReconstructionWeightsKernel.h.
Referenced by reconstruct().

PP<SelectRowsVMatrix> PLearn::ReconstructionWeightsKernel::sub_data [protected]

Points toward a subset of the training data (typically, a neighborhood).

Definition at line 114 of file ReconstructionWeightsKernel.h.
Referenced by reconstruct(), ReconstructionWeightsKernel(), and setDataForKernelMatrix().

Mat PLearn::ReconstructionWeightsKernel::weights [protected]

The matrix with the weights W_{ij}.

Definition at line 117 of file ReconstructionWeightsKernel.h.
Referenced by computeLLEMatrix(), computeWeights(), evaluate_i_j(), and evaluate_sum_k_i_k_j().

Vec PLearn::ReconstructionWeightsKernel::weights_x [mutable, private]

Used in 'evaluate_x_i_again' to store the reconstruction weights of x.

Definition at line 77 of file ReconstructionWeightsKernel.h.
Referenced by evaluate_x_i_again(), and reconstruct().

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

Generated on Tue Aug 17 16:23:41 2004 for PLearn by

1.3.7

PLearn::ReconstructionWeightsKernel Class Reference

Public Member Functions

Public Attributes

Protected Member Functions

Static Protected Member Functions

Protected Attributes

Private Types

Private Member Functions

Private Attributes

Member Typedef Documentation

Constructor & Destructor Documentation

Member Function Documentation

Member Data Documentation