PLearn 0.1
Public Types | Public Member Functions | Protected Member Functions | Protected Attributes | Friends
PLearn::TMat< T > Class Template Reference

#include <TMat_decl.h>

Inheritance diagram for PLearn::TMat< T >:
Inheritance graph
[legend]
Collaboration diagram for PLearn::TMat< T >:
Collaboration graph
[legend]

List of all members.

Public Types

typedef T value_type
typedef int size_type
typedef TMatElementIterator< T > iterator
typedef TMatElementIterator< T > const_iterator
typedef T * compact_iterator
typedef T * rowelements_iterator
typedef TMatRowsIterator< T > rows_iterator
typedef
TMatRowsAsArraysIterator< T > 
rows_as_arrays_iterator
typedef TMatColRowsIterator< T > colrows_iterator

Public Member Functions

int nrows () const
 for template compatibility with other types of matrices
int ncols () const
 TMat ()
 TMat (int the_length, int the_width)
 TMat (int the_length, int the_width, const T &init_value)
 TMat (int the_length, int the_width, T *the_data)
 TMat (int the_length, int the_width, const TVec< T > &v)
const TMat< T > & operator= (const TMat< T > &other)
 NOTE: operator= COPIES THE TMat STRUCTURE BUT NOT THE DATA (use operator<< to copy data)
iterator begin () const
 returns an iterator over elements
iterator end () const
compact_iterator compact_begin () const
 returns a compact_iterator, which is an iterator over elements, but that works only if the matrix is compact
compact_iterator compact_end () const
rowelements_iterator rowelements_begin (int rownum) const
 returns an iterator over the elements of a particular row
rowelements_iterator rowelements_end (int rownum) const
 IMPORTANT WARNING: use this only to check reaching the end with an iterator obtained through rowelements_begin USING THE *SAME* rownum.
TMatRowsIterator< T > rows_begin ()
 Return an iterator over all rows of the matrix.
TMatRowsIterator< T > rows_end ()
TMatRowsAsArraysIterator< T > rows_as_arrays_begin ()
 Return an iterator over all rows of the matrix.
TMatRowsAsArraysIterator< T > rows_as_arrays_end ()
TMatColRowsIterator< T > col_begin (int column)
 Return an iterator over a single column of the matrix.
TMatColRowsIterator< T > col_end (int column)
 This version is not strictly standards-compliant since the end-pointer is beyond 1-past-the-end-of-the-array.
void resize (int new_length, int new_width, int extra=0, bool preserve_content=false)
 Resizes the matrix to a new length() and width().
int length () const
int width () const
int size () const
int mod () const
bool isContiguous () const
 returns true if matrix elements are contiguous in memory (i.e.
bool isNotContiguous () const
 returns true if matrix elements are not contiguous in memory (i.e.
void setMod (int new_mod)
 Set a new value for 'mod'.
PP< Storage< T > > getStorage () const
bool isSquare () const
bool hasMissing () const
T * data () const
 Returns a pointer to the beginning of the matrix data.
T * operator[] (int rownum) const
 Returns a pointer to the data beginning of the required row.
T * rowdata (int i) const
T & operator() (int rownum, int colnum) const
TVec< T > operator() (int rownum) const
void write (PStream &out) const
 writes the Mat to the PStream: Note that users should rather use the form out << m;
void read (PStream &in)
 reads the Mat from the PStream: Note that users should rather use the form in >> m;
TMat< T > column (int colnum) const
 Returns a TMat that is a column of the matrix.
TMat< T > firstColumn () const
TMat< T > lastColumn () const
TMat< T > row (int row) const
 Returns a Mat that is a row of the matrix.
T & firstElement () const
T & lastElement () const
TVec< T > firstRow () const
TVec< T > lastRow () const
TVec< T > front () const
TVec< T > back () const
template<class I >
TMat< T > columns (const TVec< I > &columns) const
 selectColumns(*this,columns,result) i.e.
template<class I >
TMat< T > rows (const TVec< I > &rows) const
 selectRows(*this,rows,result) i.e.
bool operator== (const TMat< T > &other) const
bool isEqual (const TMat< T > &other, real precision=1e-6) const
template<class I >
TMat< T > operator() (const TVec< I > &rows, const TVec< I > &columns) const
TMat< T > subMat (int rowstart, int colstart, int newlength, int newwidth) const
 Returns a sub-matrix that is a rectangular portion of this matrix.
TMat< T > subMatRows (int rowstart, int newlength) const
 Returns a sub-matrix that is a range of rows of this matrix.
TMat< T > subMatColumns (int colstart, int newwidth) const
 Returns a sub-matrix that is a range of columns of this matrix.
TMat< T > copy () const
 returns a newly created copy of this Matrix
void copyTo (T *x) const
 copy to a C vector starting at x
void copyColumnTo (int col, T *x) const
 Copy a column to a C vector starting at x.
void makeDeepCopyFromShallowCopy (CopiesMap &copies)
TMat< T > deepCopy (CopiesMap &copies) const
TVec< T > toVecCopy () const
 Copy of data.
TVec< T > toVec () const
 Views same data (not always possible)
bool isNull () const
bool isNotNull () const
bool isEmpty () const
bool isNotEmpty () const
bool operator! () const
 To allow if(!m) statements.
void fill (const T &value) const
void operator= (const T &f) const
void clear () const
void swapRows (int i, int j) const
 Swap the content of row i and row j.
void swapColumns (int i, int j) const
 Swap the content of column i and column j.
int findRow (const TVec< T > &row) const
void appendRow (const TVec< T > &newrow)
void push_back (const TVec< T > &newrow)
 stl-like push_back and pop_back
void pop_back ()
void makeSharedValue (T *x, int n)
bool isCompact () const
bool isSymmetric (bool exact_check=true, bool accept_empty=false) const
 Return 'true' iff the matrix is symmetric.
void compact ()
 Ensure the allocated memory for this matrix is exactly length * width.
void transpose ()
 Swap element (i,j) with element (j,i).
void swapUpsideDown () const
void print (ostream &out=cout) const
 C++ stream output.
void input (istream &in=cin) const
void input (PStream &in) const
void debugPrint ()
void operator<< (const string &datastring) const

Protected Member Functions

void resizePreserve (int new_length, int new_width, int extra=0)
 Utility function to resize a matrix while preserving contents.
void resizeBoundCheck (int new_length, int new_width)
 Perform bound-checking on resize.
void resizeModError ()
 Report PLERROR if we resize changing the mod with usage > 1.

Protected Attributes

int offset_
int mod_
int length_
int width_
PP< Storage< T > > storage

Friends

class TVec< T >
class Variable
 for makeShared hack... (to be cleaned)
class VarArray
 for makeShared hack... (to be cleaned)

Detailed Description

template<class T>
class PLearn::TMat< T >

Definition at line 64 of file TMat_decl.h.


Member Typedef Documentation

template<class T>
typedef TMatColRowsIterator<T> PLearn::TMat< T >::colrows_iterator

Definition at line 94 of file TMat_decl.h.

template<class T>
typedef T* PLearn::TMat< T >::compact_iterator

Definition at line 89 of file TMat_decl.h.

template<class T>
typedef TMatElementIterator<T> PLearn::TMat< T >::const_iterator

Definition at line 88 of file TMat_decl.h.

template<class T>
typedef TMatElementIterator<T> PLearn::TMat< T >::iterator

Definition at line 87 of file TMat_decl.h.

template<class T>
typedef T* PLearn::TMat< T >::rowelements_iterator

Definition at line 90 of file TMat_decl.h.

Definition at line 93 of file TMat_decl.h.

template<class T>
typedef TMatRowsIterator<T> PLearn::TMat< T >::rows_iterator

Definition at line 92 of file TMat_decl.h.

template<class T>
typedef int PLearn::TMat< T >::size_type

Definition at line 86 of file TMat_decl.h.

template<class T>
typedef T PLearn::TMat< T >::value_type

Definition at line 85 of file TMat_decl.h.


Constructor & Destructor Documentation

template<class T>
PLearn::TMat< T >::TMat ( ) [inline]

Definition at line 96 of file TMat_decl.h.

        :offset_(0), mod_(0), length_(0), width_(0)
    {}
template<class T>
PLearn::TMat< T >::TMat ( int  the_length,
int  the_width 
) [inline]

Definition at line 100 of file TMat_decl.h.

        :offset_(0), mod_(0), length_(0), width_(0)
    { resize(the_length, the_width); }
template<class T>
PLearn::TMat< T >::TMat ( int  the_length,
int  the_width,
const T &  init_value 
) [inline]

Definition at line 104 of file TMat_decl.h.

        :offset_(0), mod_(0), length_(0), width_(0)
    { 
        resize(the_length, the_width); 
        fill(init_value);
    }
template<class T>
PLearn::TMat< T >::TMat ( int  the_length,
int  the_width,
T *  the_data 
) [inline]

Definition at line 111 of file TMat_decl.h.

        :offset_(0), mod_(the_width), length_(the_length), width_(the_width), 
         storage(new Storage<T>(the_length*the_width, the_data))
    {}
template<class T>
PLearn::TMat< T >::TMat ( int  the_length,
int  the_width,
const TVec< T > &  v 
)

Definition at line 282 of file TMat_impl.h.

References PLearn::TVec< T >::length(), PLearn::TMat< T >::length(), PLERROR, and PLearn::TMat< T >::width().

    : offset_(v.offset()), mod_(the_width), length_(the_length), width_(the_width), storage(v.storage)
{
    if(length()*width()!=v.length())
        PLERROR("In Mat constructor from Vec: length()*width() of matrix must be equal to length() of Vec");
}

Here is the call graph for this function:


Member Function Documentation

template<class T>
void PLearn::TMat< T >::appendRow ( const TVec< T > &  newrow) [inline]
template<class T>
TVec<T> PLearn::TMat< T >::back ( ) const [inline]

Definition at line 599 of file TMat_decl.h.

{ return lastRow(); }
template<class T >
TMatElementIterator< T > PLearn::TMat< T >::begin ( ) const [inline]
template<class T>
void PLearn::TMat< T >::clear ( ) const [inline]

Definition at line 771 of file TMat_decl.h.

Referenced by PLearn::addEigenMatrices(), PLearn::affineMatrixInitialize(), PLearn::backConvolve2D(), PLearn::backConvolve2Dbackprop(), PLearn::SparseIncrementalAffineTransformVariable::bprop(), PLearn::RBMModule::bpropAccUpdate(), PLearn::RBMConv2DConnection::bpropAccUpdate(), PLearn::OnBagsModule::bpropAccUpdate(), PLearn::NetworkModule::bpropAccUpdate(), PLearn::KLp0p1RBMModule::bpropAccUpdate(), PLearn::Convolution2DModule::bpropAccUpdate(), PLearn::RBMLateralBinomialLayer::bpropNLL(), PLearn::Supersampling2DModule::bpropUpdate(), PLearn::Subsampling2DModule::bpropUpdate(), PLearn::SquaredErrorCostModule::bpropUpdate(), PLearn::SoftmaxNLLCostModule::bpropUpdate(), PLearn::ShuntingNNetLayerModule::bpropUpdate(), PLearn::ScaleGradientModule::bpropUpdate(), PLearn::RBMWoodsLayer::bpropUpdate(), PLearn::RBMSparse1DMatrixConnection::bpropUpdate(), PLearn::RBMMultitaskClassificationModule::bpropUpdate(), PLearn::RBMMultinomialLayer::bpropUpdate(), PLearn::RBMMixedConnection::bpropUpdate(), PLearn::RBMLocalMultinomialLayer::bpropUpdate(), PLearn::RBMLateralBinomialLayer::bpropUpdate(), PLearn::RBMConv2DConnection::bpropUpdate(), PLearn::RBMBinomialLayer::bpropUpdate(), PLearn::NLLCostModule::bpropUpdate(), PLearn::LayerCostModule::bpropUpdate(), PLearn::CrossEntropyCostModule::bpropUpdate(), PLearn::Convolution2DModule::bpropUpdate(), PLearn::CombiningCostsModule::bpropUpdate(), PLearn::TreeDBNModule::build_(), PLearn::SparseIncrementalAffineTransformVariable::build_(), PLearn::OnlineGramNaturalGradientOptimizer::build_(), PLearn::MovingAverageVMatrix::build_(), PLearn::GaussianContinuumDistribution::build_(), PLearn::GaussianContinuum::build_(), PLearn::FNetLayerVariable::build_(), PLearn::EntropyContrastLearner::build_(), PLearn::DictionaryVMatrix::build_(), PLearn::RBMQLParameters::clearStats(), PLearn::RBMMatrixTransposeConnection::clearStats(), PLearn::RBMMatrixConnection::clearStats(), PLearn::RBMLQParameters::clearStats(), PLearn::RBMLLParameters::clearStats(), PLearn::RBMLateralBinomialLayer::clearStats(), PLearn::RBMGenericParameters::clearStats(), PLearn::RBMConv2DLLParameters::clearStats(), PLearn::RBMConv2DConnection::clearStats(), PLearn::LayerCostModule::computeCorrelationStatistics(), PLearn::computeCovar(), PLearn::LayerCostModule::computeHisto(), PLearn::computeInputCovar(), PLearn::ReconstructionWeightsKernel::computeLLEMatrix(), PLearn::LocalGaussianClassifier::computeOutput(), PLearn::LayerCostModule::computePascalStatistics(), PLearn::RBMSparse1DMatrixConnection::computeProducts(), PLearn::RBMMixedConnection::computeProducts(), PLearn::convolve2D(), PLearn::convolve2Dbackprop(), PLearn::correlations(), PLearn::diagonalizedFactorsTransposeProduct(), PLearn::diagonalizedFactorsTransposeProductTranspose(), PLearn::ProductRandomVariable::EMEpochInitialize(), PLearn::DenoisingRecurrentNet::encode_onehot_diffNote_duration(), PLearn::DenoisingRecurrentNet::encode_onehot_note_octav_duration(), PLearn::DenoisingRecurrentNet::encode_onehot_timeframe(), PLearn::NeuralProbabilisticLanguageModel::fillWeights(), PLearn::FeatureSetSequentialCRF::fillWeights(), PLearn::FeatureSetNNet::fillWeights(), PLearn::SubsamplingDBN::fineTuningStep(), PLearn::DeepNonLocalManifoldParzen::fineTuningStep(), PLearn::DeepBeliefNet::fineTuningStep(), PLearn::ShuntingNNetLayerModule::forget(), PLearn::RBMSparse1DMatrixConnection::forget(), PLearn::RBMQLParameters::forget(), PLearn::RBMMatrixConnection::forget(), PLearn::RBMLQParameters::forget(), PLearn::RBMLLParameters::forget(), PLearn::RBMLateralBinomialLayer::forget(), PLearn::RBMGenericParameters::forget(), PLearn::RBMConv2DLLParameters::forget(), PLearn::RBMConv2DConnection::forget(), PLearn::PseudolikelihoodRBM::forget(), PLearn::ManifoldParzen::forget(), PLearn::LayerCostModule::forget(), PLearn::GradNNetLayerModule::forget(), PLearn::ShuntingNNetLayerModule::fprop(), PLearn::RBMModule::fprop(), PLearn::ProjectionErrorVariable::fprop(), PLearn::NllSemisphericalGaussianVariable::fprop(), PLearn::NllGeneralGaussianVariable::fprop(), PLearn::LinearCombinationModule::fprop(), PLearn::LayerCostModule::fprop(), PLearn::KLp0p1RBMModule::fprop(), PLearn::RBMMixedLayer::fpropNLL(), PLearn::RBMWoodsLayer::generateSamples(), PLearn::GaussianContinuum::get_image_matrix(), PLearn::RBMSparse1DMatrixConnection::getWeights(), PLearn::OnlineGramNaturalGradientOptimizer::gramEigenNaturalGradient(), PLearn::InferenceRBM::hiddenExpGivenInput(), PLearn::NeuralProbabilisticLanguageModel::initializeParams(), PLearn::FeatureSetSequentialCRF::initializeParams(), PLearn::FeatureSetNNet::initializeParams(), PLearn::linearRegression(), PLearn::TransformationLearner::MStepTransformationDiv(), PLearn::TransformationLearner::MStepTransformations(), PLearn::SubsamplingDBN::onlineStep(), PLearn::DeepBeliefNet::onlineStep(), PLearn::NatGradEstimator::operator()(), PLearn::product(), PLearn::productTranspose(), PLearn::DenoisingRecurrentNet::recurrentUpdate(), PLearn::SparseIncrementalAffineTransformVariable::reset(), PLearn::RBMLateralBinomialLayer::reset(), PLearn::CorrelationProfiler::reset(), PLearn::smartInitialization(), PLearn::SpearmanRankCorrelation(), PLearn::PseudolikelihoodRBM::train(), PLearn::ManifoldParzen::train(), PLearn::transposeProduct(), PLearn::transposeProduct2(), PLearn::transposeTransposeProduct(), PLearn::GaussianProcessRegressor::variance(), PLearn::NeuralProbabilisticLanguageModel::verify_gradient(), PLearn::FeatureSetSequentialCRF::verify_gradient(), PLearn::FeatureSetNNet::verify_gradient(), and PLearn::weightedLinearRegression().

    { 
        if(isNotEmpty())
        {
            if(isCompact())
                clear_n(data(),size()); 
            else
            {
                int l = length();
                T* ptr = data();
                while(l--)
                {
                    clear_n(ptr, width());
                    ptr += mod();
                }
            }
        }
    }
template<class T >
TMatColRowsIterator< T > PLearn::TMat< T >::col_begin ( int  column)

Return an iterator over a single column of the matrix.

No const version for now. In other words, this iterator views a single column of the matrix AS A VECTOR to iterate on; very useful for STL algorithms.

Definition at line 513 of file TMat_impl.h.

                                                    {
    return TMatColRowsIterator<T>(data() + column, mod_);
}
template<class T >
TMatColRowsIterator< T > PLearn::TMat< T >::col_end ( int  column)

This version is not strictly standards-compliant since the end-pointer is beyond 1-past-the-end-of-the-array.

But this pointer is never dereferenced and should work on all reasonable architectures

Definition at line 518 of file TMat_impl.h.

                                                  {
    return TMatColRowsIterator<T>(data()+length_*mod_+column, mod_);
}
template<class T>
TMat<T> PLearn::TMat< T >::column ( int  colnum) const [inline]

Returns a TMat that is a column of the matrix.

Definition at line 580 of file TMat_decl.h.

Referenced by PLearn::BallTreeNearestNeighbors::anchorTrain(), PLearn::VarColumnsVariable::bprop(), PLearn::UnfoldedFuncVariable::bprop(), PLearn::LogaddOnBagsModule::bprop(), PLearn::RBMSparse1DMatrixConnection::bpropAccUpdate(), PLearn::CostModule::bpropAccUpdate(), PLearn::CombiningCostsModule::bpropAccUpdate(), PLearn::RBMMixedLayer::bpropNLL(), PLearn::RBMSparse1DMatrixConnection::bpropUpdate(), PLearn::SortRowsVMatrix::build_(), PLearn::RBMConnection::build_(), PLearn::RankedVMatrix::build_(), PLearn::MaxSubsampling2DModule::build_(), PLearn::Convolution2DModule::build_(), PLearn::StatsCollector::cdf(), PLearn::columnArgmax(), PLearn::columnArgmin(), PLearn::columnMax(), PLearn::columnMin(), PLearn::columnSumOfSquares(), PLearn::columnVariance(), PLearn::columnWeightedMean(), PLearn::columnWeightedVariance(), PLearn::AddCostToLearner::computeCostsFromOutputs(), PLearn::Kernel::computeKNNeighbourMatrixFromDistanceMatrix(), PLearn::LiftStatsCollector::computeLift(), PLearn::LiftStatsCollector::computeLiftMax(), PLearn::Kernel::computeNeighbourMatrixFromDistanceMatrix(), PLearn::BaggingLearner::computeOutput(), PLearn::RBMModule::computePartitionFunction(), PLearn::RBMSparse1DMatrixConnection::computeProducts(), PLearn::LinearRegressor::computeResidualsVariance(), PLearn::RBMQLParameters::computeUnitActivations(), PLearn::RBMLQParameters::computeUnitActivations(), PLearn::UniformizeLearner::computeWeightedRankMap(), PLearn::displayDecisionSurface(), PLearn::Kernel::estimateHistograms(), PLearn::findClosestPairsOfDifferentClass(), PLearn::SubsamplingDBN::fineTuningStep(), PLearn::DeepBeliefNet::fineTuningStep(), PLearn::UnfoldedFuncVariable::fprop(), PLearn::RBMModule::fprop(), PLearn::GaussianProcessNLLVariable::fprop(), PLearn::CostModule::fprop(), PLearn::CombiningCostsModule::fprop(), PLearn::RBMMixedLayer::fpropNLL(), PLearn::SumOverBagsVariable::fpropOneBag(), PLearn::GCV(), PLearn::VMatrixFromDistribution::getColumn(), PLearn::MemoryVMatrix::getColumn(), PLearn::SplitModule::getPortSizes(), PLearn::SubsamplingDBN::greedyStep(), PLearn::DeepBeliefNet::greedyStep(), PLearn::InferenceRBM::hiddenExpGivenInput(), PLearn::kernelPCAfromDotProducts(), PLearn::NonLocalManifoldParzen::knn(), PLearn::GaussianContinuumDistribution::knn(), PLearn::GaussianContinuum::knn(), PLearn::StatsCollector::lift(), PLearn::loadATT800(), PLearn::StatsCollector::mean_lift(), PLearn::metricMultiDimensionalScaling(), PLearn::newIndexedMatArray(), PLearn::WPLS::NIPALSEigenvector(), PLearn::PLS::NIPALSEigenvector(), PLearn::normalizeColumns(), PLearn::SubsamplingDBN::onlineStep(), PLearn::StackedAutoassociatorsNet::onlineStep(), PLearn::DeepBeliefNet::onlineStep(), PLearn::SequentialModelSelector::paired_t_test(), PLearn::StatsCollector::prbp(), PLearn::product(), PLearn::MoleculeTemplate::readFromAMATFile(), PLearn::rebalanceNClasses(), PLearn::regulargrid_x_y_outputs_to_bitmap(), PLearn::ridgeRegressionByGCV(), PLearn::TestDependencyCommand::run(), PLearn::TestDependenciesCommand::run(), PLearn::selectAndOrder(), PLearn::selectColumns(), PLearn::SpearmanRankCorrelation(), PLearn::testCholeskyRoutines(), PLearn::WPLS::train(), PLearn::StackedSVDNet::train(), PLearn::SequentialModelSelector::train(), PLearn::RankLearner::train(), PLearn::PLS::train(), PLearn::AutoLinearRegressor::train(), PLearn::RBMSparse1DMatrixConnection::update(), PLearn::GaussMix::updateSampleWeights(), and PLearn::weightedRidgeRegressionByGCV().

    { return subMatColumns(colnum, 1); }
template<class T>
template<class I >
TMat<T> PLearn::TMat< T >::columns ( const TVec< I > &  columns) const [inline]

selectColumns(*this,columns,result) i.e.

return the matrix with specified columns (indices)

Definition at line 604 of file TMat_decl.h.

    {
        TMat<T> result(length(),columns.length());
        selectColumns(*this,columns,result);
        return result;
    }
template<class T>
void PLearn::TMat< T >::compact ( ) [inline]

Ensure the allocated memory for this matrix is exactly length * width.

Definition at line 906 of file TMat_decl.h.

Referenced by PLearn::MoleculeTemplate::readFromAMATFile(), and PLearn::Molecule::readFromAMATFile().

    {
        if(storage->length() != length()*width())
        {
            if(storage->usage()>1)
                PLERROR("In TMat<T>::compact() - Compact operation not allowed"
                        " when matrix storage is shared, for obvious reasons");
            operator=(copy());
        }
    }

Here is the caller graph for this function:

template<class T>
compact_iterator PLearn::TMat< T >::compact_begin ( ) const [inline]

returns a compact_iterator, which is an iterator over elements, but that works only if the matrix is compact

Definition at line 134 of file TMat_decl.h.

Referenced by PLearn::clear(), PLearn::NeuralProbabilisticLanguageModel::gradient_penalty(), PLearn::FeatureSetSequentialCRF::gradient_penalty(), PLearn::FeatureSetNNet::gradient_penalty(), PLearn::multiply(), PLearn::multiplyAcc(), PLearn::squareElements(), PLearn::sumabs(), and PLearn::sumsquare().

    { 
#ifdef BOUNDCHECK
        if(mod()!=width()) 
            PLERROR("You cannot use a compact iterator to iterate over the elements of a non compact matrix");
#endif
        return data();
    }

Here is the caller graph for this function:

template<class T>
compact_iterator PLearn::TMat< T >::compact_end ( ) const [inline]
template<class T>
TMat<T> PLearn::TMat< T >::copy ( ) const [inline]
template<class T>
void PLearn::TMat< T >::copyColumnTo ( int  col,
T *  x 
) const [inline]

Copy a column to a C vector starting at x.

This function is thread safe.

Definition at line 696 of file TMat_decl.h.

Referenced by PLearn::RegressionTreeRegisters::getExample().

    {
        T* s = data()+col;
        for(int i=0;i<length();i++){
            *x=*s;
            x++;
            s+=mod();
        }
    }

Here is the caller graph for this function:

template<class T>
void PLearn::TMat< T >::copyTo ( T *  x) const [inline]

copy to a C vector starting at x

Definition at line 685 of file TMat_decl.h.

    {
        T* row = data(); // get data start
        int k=0;
        for(int i=0; i<length(); i++,row+=mod())
            for (int j=0;j<width();j++,k++)
                x[k] = row[j];
    }
template<class T>
T* PLearn::TMat< T >::data ( ) const [inline]

Returns a pointer to the beginning of the matrix data.

Definition at line 324 of file TMat_decl.h.

Referenced by PLearn::absargmax(), PLearn::add(), PLearn::addToColumns(), PLearn::addToDiagonal(), PLearn::argmax(), PLearn::argmin(), PLearn::averageAcrossRowsAndColumns(), PLearn::backConvolve2D(), PLearn::backConvolve2Dbackprop(), PLearn::SparseIncrementalAffineTransformVariable::bprop(), PLearn::SoftSoftMaxVariable::bprop(), PLearn::MaxSubsampling2DModule::bpropAccUpdate(), PLearn::RBMLateralBinomialLayer::bpropNLL(), PLearn::RBMMultinomialLayer::bpropUpdate(), PLearn::RBMLateralBinomialLayer::bpropUpdate(), PLearn::chol_rotapp_tr_opt(), PLearn::choleskyDecomposition(), PLearn::choleskyInvert(), PLearn::choleskySolve(), PLearn::RationalQuadraticARDKernel::computeGramMatrix(), PLearn::RationalQuadraticARDKernel::computeGramMatrixDerivIspAlpha(), PLearn::SquaredExponentialARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::RationalQuadraticARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::NeuralNetworkARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::Matern1ARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::LinearARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::computeInverseStandardDeviationFromMeanAndSquareMean(), PLearn::convolve(), PLearn::convolve2D(), PLearn::convolve2Dbackprop(), PLearn::det(), PLearn::diagonalizedFactorsProduct(), PLearn::diagonalizedFactorsProductBprop(), PLearn::diagonalizedFactorsProductTranspose(), PLearn::diagonalizedFactorsProductTransposeBprop(), PLearn::diagonalizedFactorsTransposeProduct(), PLearn::diagonalizedFactorsTransposeProductBprop(), PLearn::diagonalizedFactorsTransposeProductTranspose(), PLearn::diagonalizedFactorsTransposeProductTransposeBprop(), PLearn::diffSquareMultiplyAcc(), PLearn::dot(), PLearn::eigen_SymmMat(), PLearn::eigenSparseNonSymmMat(), PLearn::eigenSparseSymmMat(), PLearn::ProductRandomVariable::EMBprop(), PLearn::externalProductAcc(), PLearn::RBMLateralBinomialLayer::externalSymetricProductAcc(), PLearn::RBMLateralBinomialLayer::forget(), PLearn::SoftSoftMaxVariable::fprop(), PLearn::geometric_mean(), PLearn::DenoisingRecurrentNet::inject_zero_forcing_noise(), PLearn::inverse(), PLearn::invertElements(), PLearn::lapackCholeskyDecompositionInPlace(), PLearn::lapackCholeskySolveInPlace(), PLearn::lapackEIGEN(), PLearn::lapackGeneralizedEIGEN(), PLearn::lapackSolveLinearSystem(), PLearn::lapackSVD(), PLearn::linearRegression(), PLearn::loadADMat(), PLearn::loadCorelDatamat(), PLearn::loadJPEGrgb(), PLearn::loadSTATLOG(), PLearn::loadUCIMLDB(), PLearn::LU_decomposition(), PLearn::matColumnDotVec(), PLearn::matInvert(), PLearn::max(), PLearn::maxabs(), PLearn::mean(), PLearn::min(), PLearn::minabs(), PLearn::multiplyAcc(), PLearn::multiplyScaledAdd(), PLearn::negateElements(), PLearn::ConvertToPyObject< Mat >::newPyObject(), PLearn::StackedAutoassociatorsNet::onlineStep(), PLearn::operator*=(), PLearn::operator+=(), PLearn::operator-(), PLearn::operator-=(), PLearn::operator/=(), PLearn::partialSortRows(), PLearn::RBMMatrixConnection::petiteCulotteOlivierCD(), PLearn::product(), PLearn::product2Acc(), PLearn::productAcc(), PLearn::productScaleAcc(), PLearn::qld_interface(), PLearn::regularizeMatrix(), PLearn::sortRows(), PLearn::squareMultiplyAcc(), PLearn::squareProductAcc(), PLearn::substract(), PLearn::substractFromColumns(), PLearn::sum(), PLearn::sum_of_squares(), PLearn::transposeProduct(), PLearn::transposeProduct2(), PLearn::transposeProduct2Acc(), PLearn::transposeProductAcc(), PLearn::transposeProductScaleAcc(), PLearn::transposeTransposeProduct(), PLearn::transposeTransposeProductAcc(), PLearn::transposeTransposeProductScaleAcc(), PLearn::RBMMatrixTransposeConnection::update(), PLearn::RBMMatrixConnection::update(), PLearn::RBMLLParameters::update(), PLearn::RBMLateralBinomialLayer::update(), PLearn::RBMConv2DLLParameters::update(), PLearn::RBMConv2DConnection::update(), PLearn::NeuralProbabilisticLanguageModel::update_affine_transform(), PLearn::FeatureSetSequentialCRF::update_affine_transform(), PLearn::FeatureSetNNet::update_affine_transform(), and PLearn::variance().

    {
#ifdef BOUNDCHECK
        if(storage.isNull())
            PLERROR("IN TMat::data()\nAttempted to get a pointer to the data of an empty matrix");
#endif
        return storage->data+offset_; 
    }
template<class T>
void PLearn::TMat< T >::debugPrint ( ) [inline]

Definition at line 946 of file TMat_decl.h.

{print(cerr);}
template<class T >
TMat< T > PLearn::TMat< T >::deepCopy ( CopiesMap copies) const

Notice that deepCopy of a Mat returns a Mat rather than a Mat*. The reason for this being that a Mat is already some kind of "smart pointer" to an underlying Storage

Definition at line 470 of file TMat_impl.h.

References PLearn::TMat< T >::makeDeepCopyFromShallowCopy().

Referenced by PLearn::deepCopy().

{
    // First do a shallow copy
    TMat<T> deep_copy = *this;
    // Transform the shallow copy into a deep copy
    deep_copy.makeDeepCopyFromShallowCopy(copies);
    // return the completed deep_copy
    return deep_copy;
}

Here is the call graph for this function:

Here is the caller graph for this function:

template<class T >
TMatElementIterator< T > PLearn::TMat< T >::end ( ) const [inline]
template<class T>
void PLearn::TMat< T >::fill ( const T &  value) const [inline]

Definition at line 750 of file TMat_decl.h.

Referenced by align_(), PLearn::VecStatsCollector::append(), PLearn::Supersampling2DModule::bbpropUpdate(), PLearn::Subsampling2DModule::bbpropUpdate(), PLearn::Supersampling2DModule::bpropUpdate(), PLearn::Subsampling2DModule::bpropUpdate(), PLearn::SoftmaxModule::bpropUpdate(), PLearn::ShuntingNNetLayerModule::bpropUpdate(), PLearn::OnlineLearningModule::bpropUpdate(), PLearn::OnBagsModule::bpropUpdate(), PLearn::LinearFilterModule::bpropUpdate(), PLearn::GradNNetLayerModule::bpropUpdate(), PLearn::SplitWiseValidationVMatrix::build_(), PLearn::SequentialLearner::build_(), PLearn::RBMModule::build_(), PLearn::RBMConnection::build_(), PLearn::OnBagsModule::build_(), PLearn::MovingAverageVMatrix::build_(), PLearn::ModuleTester::build_(), PLearn::MaxSubsampling2DModule::build_(), PLearn::LayerCostModule::build_(), PLearn::KLp0p1RBMModule::build_(), PLearn::FNetLayerVariable::build_(), PLearn::CumVMatrix::build_(), PLearn::Convolution2DModule::build_(), PLearn::BackConvolution2DModule::build_(), PLearn::ChemicalICP::cacheFeatureDistances(), PLearn::DeepReconstructorNet::computeAndSaveLayerActivationStats(), PLearn::LayerCostModule::computeCorrelationStatistics(), PLearn::PLearnerDiagonalKernel::computeGramMatrix(), PLearn::IIDNoiseKernel::computeGramMatrixDerivative(), PLearn::GaussMix::computeLogLikelihood(), PLearn::LocalGaussianClassifier::computeOutput(), PLearn::KNNClassifier::computeOutput(), PLearn::GaussianProcessRegressor::computeOutputCovMat(), PLearn::ModuleLearner::computeOutputsAndCosts(), PLearn::DeepBeliefNet::computeOutputsAndCosts(), PLearn::computeRanks(), PLearn::LayerCostModule::computeSafeHisto(), PLearn::correlations(), PLearn::displayVarGraph(), PLearn::MemoryVMatrix::fill(), PLearn::fill_one_hot(), PLearn::QuantilesStatsIterator::finish(), PLearn::SequentialLearner::forget(), PLearn::IncrementalNNet::forget(), PLearn::Supersampling2DModule::fprop(), PLearn::Subsampling2DModule::fprop(), PLearn::RBMModule::fprop(), PLearn::LayerCostModule::fprop(), PLearn::SplitModule::getPortSizes(), PLearn::OnlineLearningModule::getPortSizes(), PLearn::NullModule::getPortSizes(), PLearn::CostModule::getPortSizes(), PLearn::BinarizeModule::getPortSizes(), PLearn::InferenceRBM::hiddenExpGivenInput(), PLearn::loadCorel(), PLearn::LocallyMagnifiedDistribution::log_density(), PLearn::VecStatsCollector::merge(), PLearn::TransformationLearner::MStepBias(), PLearn::OldDisplayVarGraph(), PLearn::TMatTest::perform(), PLearn::SourceVariable::randomInitialize(), PLearn::NnlmOutputLayer::resetAllClassVars(), PLearn::NnlmOutputLayer::resetParameters(), PLearn::UndirectedSoftmaxModule::resetWeights(), PLearn::NnlmWordRepresentationLayer::resetWeights(), PLearn::EntropyContrast::set_NNcontinuous_gradient_from_extra_cost(), PLearn::GeodesicDistanceKernel::setDataForKernelMatrix(), PLearn::MatrixModule::setGradientTo(), PLearn::GaussMix::setPredictor(), PLearn::GaussMix::setPredictorPredictedSizes_const(), PLearn::StackedFocusedAutoassociatorsNet::setTrainingSet(), PLearn::RowMapSparseMatrix< real >::toMat(), PLearn::WPLS::train(), PLearn::VariableSelectionWithDirectedGradientDescent::train(), PLearn::SubsamplingDBN::train(), PLearn::PLS::train(), PLearn::NatGradSMPNNet::train(), PLearn::NatGradNNet::train(), PLearn::mNNet::train(), PLearn::LinearInductiveTransferClassifier::train(), PLearn::GaussMix::train(), PLearn::DeepBeliefNet::train(), PLearn::VecStatsCollector::update(), and PLearn::GaussMix::updateInverseVarianceFromPrevious().

    {
        if (isNotEmpty()) {
            if(isCompact())
                fill_n(data(),size(),value); 
            else
            {
                int l = length();
                T* ptr = data();
                while(l--)
                {
                    fill_n(ptr, width(), value);
                    ptr += mod();
                }
            }
        }
    }
template<class T>
int PLearn::TMat< T >::findRow ( const TVec< T > &  row) const

Definition at line 315 of file TMat_impl.h.

References i.

{
    for(int i=0; i<length(); i++)
        if( (*this)(i)==row )
            return i;
    return -1;
}
template<class T>
TMat<T> PLearn::TMat< T >::firstColumn ( ) const [inline]

Definition at line 583 of file TMat_decl.h.

    { return column(0); }
template<class T>
T& PLearn::TMat< T >::firstElement ( ) const [inline]

Definition at line 593 of file TMat_decl.h.

Referenced by PLearn::trace().

{ return *data(); }

Here is the caller graph for this function:

template<class T>
TVec<T> PLearn::TMat< T >::firstRow ( ) const [inline]
template<class T>
TVec<T> PLearn::TMat< T >::front ( ) const [inline]

Definition at line 598 of file TMat_decl.h.

{ return firstRow(); }
template<class T>
PP< Storage<T> > PLearn::TMat< T >::getStorage ( ) const [inline]

Definition at line 304 of file TMat_decl.h.

    { return storage; }
template<class T>
bool PLearn::TMat< T >::hasMissing ( ) const [inline]
template<class T >
void PLearn::TMat< T >::input ( PStream in) const

Definition at line 370 of file TMat_impl.h.

References i, j, and PLERROR.

{
    for(int i=0; i<length(); i++)
    {
        T* v = rowdata(i);
        for (int j=0;j<width();j++)
        {
            if (!in)
                PLERROR("In TMat<T>::input error encountered while reading matrix");
            else
                in>>v[j];
        }
    }
}
template<class T >
void PLearn::TMat< T >::input ( istream &  in = cin) const

Definition at line 356 of file TMat_impl.h.

References i, j, and PLERROR.

Referenced by PLearn::operator>>().

{
    for(int i=0; i<length(); i++)
    {
        T* v = rowdata(i);
        for (int j=0;j<width();j++)
        {
            if(!(in>>v[j]))
                PLERROR("In TMat<T>::input error encountered while reading matrix");
        }
    }
}

Here is the caller graph for this function:

template<class T>
bool PLearn::TMat< T >::isCompact ( ) const [inline]
template<class T>
bool PLearn::TMat< T >::isContiguous ( ) const [inline]

returns true if matrix elements are contiguous in memory (i.e.

no gap between last element of a row and first element of next row).

Definition at line 259 of file TMat_decl.h.

    { return mod_==width_; }
template<class T>
bool PLearn::TMat< T >::isEmpty ( ) const [inline]

Definition at line 731 of file TMat_decl.h.

Referenced by PLearn::TreeDBNModule::bpropAccUpdate(), PLearn::SoftmaxNLLCostModule::bpropAccUpdate(), PLearn::RBMSparse1DMatrixConnection::bpropAccUpdate(), PLearn::RBMModule::bpropAccUpdate(), PLearn::RBMMixedConnection::bpropAccUpdate(), PLearn::RBMMatrixConnection::bpropAccUpdate(), PLearn::RBMConv2DConnection::bpropAccUpdate(), PLearn::OnlineLearningModule::bpropAccUpdate(), PLearn::OnBagsModule::bpropAccUpdate(), PLearn::NLLCostModule::bpropAccUpdate(), PLearn::NetworkModule::bpropAccUpdate(), PLearn::MaxSubsampling2DModule::bpropAccUpdate(), PLearn::MatrixModule::bpropAccUpdate(), PLearn::LinearCombinationModule::bpropAccUpdate(), PLearn::LayerCostModule::bpropAccUpdate(), PLearn::KLp0p1RBMModule::bpropAccUpdate(), PLearn::IdentityModule::bpropAccUpdate(), PLearn::CrossEntropyCostModule::bpropAccUpdate(), PLearn::CostModule::bpropAccUpdate(), PLearn::Convolution2DModule::bpropAccUpdate(), PLearn::OnlineLearningModule::bpropUpdate(), PLearn::RemapLastColumnVMatrix::build_(), PLearn::ClassErrorCostModule::build_(), PLearn::RBMModule::computeFreeEnergyOfHidden(), PLearn::KLp0p1RBMModule::computeFreeEnergyOfHidden(), PLearn::RBMModule::computeFreeEnergyOfVisible(), PLearn::KLp0p1RBMModule::computeFreeEnergyOfVisible(), PLearn::RBMModule::computeHiddenActivations(), PLearn::KLp0p1RBMModule::computeHiddenActivations(), PLearn::RBMModule::computePositivePhaseHiddenActivations(), PLearn::KLp0p1RBMModule::computePositivePhaseHiddenActivations(), PLearn::RBMModule::computeVisibleActivations(), PLearn::KLp0p1RBMModule::computeVisibleActivations(), PLearn::eigenVecOfSymmMat(), PLearn::VBoundDBN2::fprop(), PLearn::TreeDBNModule::fprop(), PLearn::SquaredErrorCostModule::fprop(), PLearn::SoftmaxNLLCostModule::fprop(), PLearn::RBMModule::fprop(), PLearn::OnlineLearningModule::fprop(), PLearn::OnBagsModule::fprop(), PLearn::NLLCostModule::fprop(), PLearn::NetworkModule::fprop(), PLearn::MaxSubsampling2DModule::fprop(), PLearn::LinearCombinationModule::fprop(), PLearn::LayerCostModule::fprop(), PLearn::KLp0p1RBMModule::fprop(), PLearn::CostModule::fprop(), PLearn::Convolution2DModule::fprop(), PLearn::ClassErrorCostModule::fprop(), PLearn::ArgmaxModule::fprop(), PLearn::TreeDBNModule::full_fprop(), PLearn::RemapLastColumnVMatrix::getNewRow(), PLearn::SplitModule::getPortSizes(), PLearn::NullModule::getPortSizes(), PLearn::lapackEIGEN(), PLearn::lapackGeneralizedEIGEN(), PLearn::logadd(), PLearn::matInvert(), PLearn::ConvertToPyObject< Mat >::newPyObject(), PLearn::product(), PLearn::productAcc(), PLearn::productScaleAcc(), PLearn::productTranspose(), PLearn::productTransposeAcc(), PLearn::productTransposeScaleAcc(), PLearn::transposeProduct(), PLearn::transposeProductAcc(), PLearn::transposeProductScaleAcc(), PLearn::transposeTransposeProduct(), PLearn::transposeTransposeProductAcc(), and PLearn::transposeTransposeProductScaleAcc().

    { return length_ == 0 || width_ == 0; }
template<class T>
bool PLearn::TMat< T >::isEqual ( const TMat< T > &  other,
real  precision = 1e-6 
) const [inline]

Definition at line 540 of file TMat_impl.h.

References PLearn::TMat< T >::begin(), PLearn::is_equal(), PLearn::TMat< T >::length(), and PLearn::TMat< T >::width().

Referenced by PLearn::save_load_compare().

{
    if ( length() != other.length() || width() != other.width() )
        return false;
  
    iterator it       = begin();
    iterator end_     = end();
    iterator other_it = other.begin();

    for(; it != end_; ++it, ++other_it)
        if( !is_equal(*it,*other_it, 10.0, precision, precision) )
            return false;
  
    return true;
}

Here is the call graph for this function:

Here is the caller graph for this function:

template<class T>
bool PLearn::TMat< T >::isNotContiguous ( ) const [inline]

returns true if matrix elements are not contiguous in memory (i.e.

there is a gap between last element of a row and first element of next row).

Definition at line 264 of file TMat_decl.h.

Referenced by PLearn::UnaryVariable::checkContiguity().

    { return mod_!=width_; }

Here is the caller graph for this function:

template<class T>
bool PLearn::TMat< T >::isNotEmpty ( ) const [inline]
template<class T>
bool PLearn::TMat< T >::isNotNull ( ) const [inline]
template<class T>
bool PLearn::TMat< T >::isNull ( ) const [inline]

Definition at line 725 of file TMat_decl.h.

Referenced by PLearn::ICP::buildMeshes(), and PLearn::ConvertToPyObject< Mat >::newPyObject().

    { return storage.isNull(); }

Here is the caller graph for this function:

template<class T>
bool PLearn::TMat< T >::isSquare ( ) const [inline]

Definition at line 307 of file TMat_decl.h.

Referenced by PLearn::makeItSymmetric(), and PLearn::trace().

    { return length() == width(); }

Here is the caller graph for this function:

template<class T>
bool PLearn::TMat< T >::isSymmetric ( bool  exact_check = true,
bool  accept_empty = false 
) const [inline]

Return 'true' iff the matrix is symmetric.

If 'exact_check' is true, it performs a fast exact equality check (which does not handle 'nan' or 'inf' for instance), otherwise it uses the approximate and slower 'is_equal' function from pl_math.h. If 'accept_empty' is set to 'true', then empty matrices will be considered as symmetric, otherwise a warning will be issued and 'false' will be returned.

Definition at line 870 of file TMat_decl.h.

Referenced by PLearn::GaussMix::addToCovariance(), PLearn::GaussMix::computeLogLikelihood(), PLearn::GaussMix::computeMeansAndCovariances(), PLearn::eigen_SymmMat(), PLearn::eigenVecOfSymmMat(), PLearn::GaussianProcessNLLVariable::fbpropFragments(), PLearn::GaussMix::setPredictor(), PLearn::GaussMix::setPredictorPredictedSizes_const(), and PLearn::GaussMix::updateInverseVarianceFromPrevious().

    {
        if (!isSquare())
            return false;

        if (isEmpty())
        {
            if (accept_empty)
                return true;
            else {
                PLWARNING("In TMat::isSymmetric - The matrix is empty, considering "
                          "it is not symmetric (use 'accept_empty' if you want to "
                          " allow it)");
                return false;
            }
        }

        int n = length();
        PLASSERT( width() == n );

        if (exact_check) {
            for (int i = 0; i < n - 1 ; i++)
                for (int j = i + 1; j < n; j++)
                    if ( !fast_exact_is_equal((*this)[i][j], (*this)[j][i]) )
                        return false;
        } else {
            for (int i = 0; i < n ; i++)
                for (int j = i + 1; j < n; j++)
                    if ( !is_equal((*this)[i][j], (*this)[j][i] ) )
                        return false;
        }

        return true;
    }

Here is the caller graph for this function:

template<class T>
TMat<T> PLearn::TMat< T >::lastColumn ( ) const [inline]
template<class T>
T& PLearn::TMat< T >::lastElement ( ) const [inline]

Definition at line 594 of file TMat_decl.h.

Referenced by PLearn::DTWKernel::evaluate().

{ return operator()(length_-1,width_-1); }

Here is the caller graph for this function:

template<class T>
TVec<T> PLearn::TMat< T >::lastRow ( ) const [inline]

Definition at line 597 of file TMat_decl.h.

{ return operator()(length_ - 1); }
template<class T>
int PLearn::TMat< T >::length ( ) const [inline]

Definition at line 245 of file TMat_decl.h.

Referenced by PLearn::absargmax(), PLearn::RBMSparse1DMatrixConnection::accumulateNegStats(), PLearn::RBMMatrixConnection::accumulateNegStats(), PLearn::RBMLayer::accumulateNegStats(), PLearn::RBMDiagonalMatrixConnection::accumulateNegStats(), PLearn::RBMSparse1DMatrixConnection::accumulatePosStats(), PLearn::RBMMatrixConnection::accumulatePosStats(), PLearn::RBMLayer::accumulatePosStats(), PLearn::RBMDiagonalMatrixConnection::accumulatePosStats(), PLearn::add(), PLearn::RBMLayer::addBiasDecay(), PLearn::MemoryCachedKernel::addDataForKernelMatrix(), PLearn::PLearnDiff::addDiffPrefix(), PLearn::addEigenMatrices(), PLearn::addToColumns(), PLearn::GaussMix::addToCovariance(), PLearn::addToDiagonal(), PLearn::addToRows(), PLearn::RBMMatrixConnection::addWeightPenalty(), PLearn::affineMatrixInitialize(), align(), align_(), PLearn::GaussianProcessNLLVariable::alpha(), PLearn::VecStatsCollector::append(), PLearn::BasisSelectionRegressor::appendKernelFunctions(), PLearn::VMatrix::appendRows(), PLearn::apply(), PLearn::Learner::applyAndComputeCostsOnTestMat(), PLearn::applyGeomTransformation(), PLearn::argmax(), PLearn::argmin(), PLearn::averageAcrossRowsAndColumns(), PLearn::backConvolve2D(), PLearn::backConvolve2Dbackprop(), PLearn::TransformationLearner::batchGeneratePredictedFrom(), PLearn::GaussianProcessRegressor::BayesianCost(), PLearn::binary_search(), PLearn::bootstrap_rows(), PLearn::WeightedLogGaussian::bprop(), PLearn::TransposedDoubleProductVariable::bprop(), PLearn::TraceVariable::bprop(), PLearn::SumVarianceOfLinearTransformedCategoricals::bprop(), PLearn::SumVarianceOfLinearTransformedBernoullis::bprop(), PLearn::SumEntropyOfCategoricals::bprop(), PLearn::SumEntropyOfBernoullis::bprop(), PLearn::SparseIncrementalAffineTransformVariable::bprop(), PLearn::NonDiagVariable::bprop(), PLearn::MultiMaxVariable::bprop(), PLearn::MatrixAffineTransformVariable::bprop(), PLearn::LogSoftSoftMaxVariable::bprop(), PLearn::LogaddOnBagsModule::bprop(), PLearn::LocalizedFeaturesLayerVariable::bprop(), PLearn::InsertZerosVariable::bprop(), PLearn::GaussianProcessNLLVariable::bprop(), PLearn::DoubleProductVariable::bprop(), PLearn::Cov2CorrVariable::bprop(), PLearn::ConstrainVariable::bprop(), PLearn::ConcatColumnsVariable::bprop(), PLearn::AffineTransformVariable::bprop(), PLearn::NllGeneralGaussianVariable::bprop_to_bases(), PLearn::DeepNonLocalManifoldParzen::bprop_to_bases(), PLearn::VBoundDBN2::bpropAccUpdate(), PLearn::TreeDBNModule::bpropAccUpdate(), PLearn::SplitModule::bpropAccUpdate(), PLearn::SoftmaxNLLCostModule::bpropAccUpdate(), PLearn::RBMSparse1DMatrixConnection::bpropAccUpdate(), PLearn::RBMModule::bpropAccUpdate(), PLearn::RBMMixedConnection::bpropAccUpdate(), PLearn::RBMMatrixConnection::bpropAccUpdate(), PLearn::RBMConv2DConnection::bpropAccUpdate(), PLearn::OnlineLearningModule::bpropAccUpdate(), PLearn::OnBagsModule::bpropAccUpdate(), PLearn::NLLCostModule::bpropAccUpdate(), PLearn::NetworkModule::bpropAccUpdate(), PLearn::MaxSubsampling2DModule::bpropAccUpdate(), PLearn::MatrixModule::bpropAccUpdate(), PLearn::LinearCombinationModule::bpropAccUpdate(), PLearn::LayerCostModule::bpropAccUpdate(), PLearn::KLp0p1RBMModule::bpropAccUpdate(), PLearn::IdentityModule::bpropAccUpdate(), PLearn::CrossEntropyCostModule::bpropAccUpdate(), PLearn::CostModule::bpropAccUpdate(), PLearn::Convolution2DModule::bpropAccUpdate(), PLearn::CombiningCostsModule::bpropAccUpdate(), PLearn::BinarizeModule::bpropAccUpdate(), PLearn::bpropCholeskyDecomposition(), PLearn::bpropCholeskySolve(), PLearn::mNNet::bpropNet(), PLearn::RBMWoodsLayer::bpropNLL(), PLearn::RBMMultinomialLayer::bpropNLL(), PLearn::RBMMixedLayer::bpropNLL(), PLearn::RBMLocalMultinomialLayer::bpropNLL(), PLearn::RBMLateralBinomialLayer::bpropNLL(), PLearn::RBMGaussianLayer::bpropNLL(), PLearn::RBMBinomialLayer::bpropNLL(), PLearn::TanhModule::bpropUpdate(), PLearn::SquaredErrorCostModule::bpropUpdate(), PLearn::SoftmaxNLLCostModule::bpropUpdate(), PLearn::SoftmaxModule::bpropUpdate(), PLearn::ShuntingNNetLayerModule::bpropUpdate(), PLearn::ScaleGradientModule::bpropUpdate(), PLearn::RBMWoodsLayer::bpropUpdate(), PLearn::RBMSparse1DMatrixConnection::bpropUpdate(), PLearn::RBMMultinomialLayer::bpropUpdate(), PLearn::RBMMixedLayer::bpropUpdate(), PLearn::RBMMixedConnection::bpropUpdate(), PLearn::RBMMatrixTransposeConnection::bpropUpdate(), PLearn::RBMMatrixConnectionNatGrad::bpropUpdate(), PLearn::RBMMatrixConnection::bpropUpdate(), PLearn::RBMLocalMultinomialLayer::bpropUpdate(), PLearn::RBMLateralBinomialLayer::bpropUpdate(), PLearn::RBMDiagonalMatrixConnection::bpropUpdate(), PLearn::RBMConv2DConnection::bpropUpdate(), PLearn::RBMBinomialLayer::bpropUpdate(), PLearn::ProcessInputCostModule::bpropUpdate(), PLearn::OnlineLearningModule::bpropUpdate(), PLearn::OnBagsModule::bpropUpdate(), PLearn::NLLCostModule::bpropUpdate(), PLearn::ModuleStackModule::bpropUpdate(), PLearn::LinearFilterModule::bpropUpdate(), PLearn::LayerCostModule::bpropUpdate(), PLearn::GradNNetLayerModule::bpropUpdate(), PLearn::CrossEntropyCostModule::bpropUpdate(), PLearn::CombiningCostsModule::bpropUpdate(), PLearn::mNNet::bpropUpdateNet(), PLearn::YMDDatedVMatrix::build_(), PLearn::SplitWiseValidationVMatrix::build_(), PLearn::ShuntingNNetLayerModule::build_(), PLearn::ScoreLayerVariable::build_(), PLearn::RepeatSplitter::build_(), PLearn::RBMSparse1DMatrixConnection::build_(), PLearn::RBMQLParameters::build_(), PLearn::RBMMatrixConnection::build_(), PLearn::RBMLQParameters::build_(), PLearn::RBMLLParameters::build_(), PLearn::RBMGenericParameters::build_(), PLearn::RBMConv2DLLParameters::build_(), PLearn::RBMConv2DConnection::build_(), PLearn::RankedVMatrix::build_(), PLearn::PLearnDiff::build_(), PLearn::ObjectOptionVariable::build_(), PLearn::MultiTargetOneHotVMatrix::build_(), PLearn::MoleculeTemplate::build_(), PLearn::ModuleTester::build_(), PLearn::MemoryVMatrix::build_(), PLearn::LocalNeighborsDifferencesVMatrix::build_(), PLearn::LinearInductiveTransferClassifier::build_(), PLearn::LemmatizeVMatrix::build_(), PLearn::KNNImputationVMatrix::build_(), PLearn::GradNNetLayerModule::build_(), PLearn::FNetLayerVariable::build_(), PLearn::ExtractNNetParamsVMatrix::build_(), PLearn::DictionaryVMatrix::build_(), PLearn::Convolution2DModule::build_(), PLearn::ClassErrorCostModule::build_(), PLearn::BackConvolution2DModule::build_(), PLearn::AppendNeighborsVMatrix::build_(), PLearn::Convolution2DModule::build_kernels(), PLearn::BackConvolution2DModule::build_kernels(), PLearn::BasisSelectionRegressor::buildTopCandidateFunctions(), PLearn::ChemicalICP::cacheFeatureDistances(), PLearn::RBMTrainer::CD1(), PLearn::chol_dxch(), PLearn::chol_rotapp_tr(), PLearn::choleskyAppendDimension(), PLearn::choleskyDecomposition(), PLearn::choleskyInsertBasis(), PLearn::choleskyInvert(), PLearn::choleskyLeftSolve(), PLearn::choleskyRemoveDimension(), PLearn::choleskyRightSolve(), PLearn::choleskySolve(), PLearn::choleskyUpgrade(), PLearn::classification_confusion_matrix(), PLearn::columnArgmax(), PLearn::columnArgmin(), PLearn::columnMax(), PLearn::columnMean(), PLearn::columnMin(), PLearn::columnSum(), PLearn::columnVariance(), PLearn::columnWeightedMean(), PLearn::columnWeightedVariance(), PLearn::RBMModule::computeAllHiddenProbabilities(), PLearn::DeepBeliefNet::computeClassifAndFinalCostsFromOutputs(), PLearn::computeColumnsMeanAndStddev(), PLearn::computeConditionalMeans(), PLearn::LayerCostModule::computeCorrelationStatistics(), PLearn::AddCostToLearner::computeCostsFromOutputs(), PLearn::computeCovar(), PLearn::RBMModule::computeEnergy(), PLearn::KLp0p1RBMModule::computeEnergy(), PLearn::RBMWoodsLayer::computeExpectations(), PLearn::RBMTruncExpLayer::computeExpectations(), PLearn::RBMRateLayer::computeExpectations(), PLearn::RBMMultinomialLayer::computeExpectations(), PLearn::RBMLocalMultinomialLayer::computeExpectations(), PLearn::RBMLateralBinomialLayer::computeExpectations(), PLearn::RBMGaussianLayer::computeExpectations(), PLearn::RBMBinomialLayer::computeExpectations(), PLearn::RBMModule::computeFreeEnergyOfHidden(), PLearn::KLp0p1RBMModule::computeFreeEnergyOfHidden(), PLearn::RBMModule::computeFreeEnergyOfVisible(), PLearn::KLp0p1RBMModule::computeFreeEnergyOfVisible(), PLearn::ThresholdedKernel::computeGramMatrix(), PLearn::SummationKernel::computeGramMatrix(), PLearn::RationalQuadraticARDKernel::computeGramMatrix(), PLearn::PLearnerDiagonalKernel::computeGramMatrix(), PLearn::KroneckerBaseKernel::computeGramMatrix(), PLearn::Kernel::computeGramMatrix(), PLearn::IIDNoiseKernel::computeGramMatrix(), PLearn::RationalQuadraticARDKernel::computeGramMatrixDerivIspAlpha(), PLearn::Matern1ARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::RationalQuadraticARDKernel::computeGramMatrixDerivIspSignalSigma(), PLearn::PLearnerDiagonalKernel::computeGramMatrixDerivIspSignalSigma(), PLearn::Matern1ARDKernel::computeGramMatrixDerivIspSignalSigma(), PLearn::IIDNoiseKernel::computeGramMatrixDerivKronecker(), PLearn::MemoryCachedKernel::computeGramMatrixNV(), PLearn::RBMModule::computeHiddenActivations(), PLearn::KLp0p1RBMModule::computeHiddenActivations(), PLearn::LayerCostModule::computeHisto(), PLearn::RGBImageDB::computeHistogramRepresentation(), PLearn::LinearRegressor::computeInformationCriteria(), PLearn::computeInverseStandardDeviationFromMeanAndSquareMean(), PLearn::LayerCostModule::computeKLdiv(), PLearn::Kernel::computeKNNeighbourMatrixFromDistanceMatrix(), PLearn::ReconstructionWeightsKernel::computeLLEMatrix(), PLearn::GaussMix::computeLogLikelihood(), PLearn::RBMModule::computeLogLikelihoodOfVisible(), PLearn::DeepNonLocalManifoldParzen::computeManifoldParzenParameters(), PLearn::computeMeanAndCovar(), PLearn::computeNearestNeighbors(), PLearn::RBMModule::computeNegLogPVisibleGivenPHidden(), PLearn::Kernel::computeNeighbourMatrixFromDistanceMatrix(), PLearn::ManifoldParzen2::computeOutput(), PLearn::KernelRidgeRegressor::computeOutput(), PLearn::IsomapTangentLearner::computeOutput(), PLearn::BaggingLearner::computeOutput(), PLearn::ModuleLearner::computeOutputAndCosts(), PLearn::PLearner::computeOutputCovMat(), PLearn::GaussianProcessRegressor::computeOutputCovMat(), PLearn::StackedAutoassociatorsNet::computeOutputs(), PLearn::PLearner::computeOutputs(), PLearn::NatGradNNet::computeOutputs(), PLearn::mNNet::computeOutputs(), PLearn::StackedAutoassociatorsNet::computeOutputsAndCosts(), PLearn::PLearner::computeOutputsAndCosts(), PLearn::NatGradNNet::computeOutputsAndCosts(), PLearn::ModuleLearner::computeOutputsAndCosts(), PLearn::mNNet::computeOutputsAndCosts(), PLearn::DeepBeliefNet::computeOutputsAndCosts(), PLearn::AddCostToLearner::computeOutputsAndCosts(), PLearn::Kernel::computePartialGramMatrix(), PLearn::LayerCostModule::computePascalStatistics(), PLearn::RBMModule::computePositivePhaseHiddenActivations(), PLearn::KLp0p1RBMModule::computePositivePhaseHiddenActivations(), PLearn::GaussMix::computePosteriors(), PLearn::computePrincipalComponents(), PLearn::RBMSparse1DMatrixConnection::computeProducts(), PLearn::RBMMatrixTransposeConnection::computeProducts(), PLearn::RBMMatrixConnection::computeProducts(), PLearn::RBMDiagonalMatrixConnection::computeProducts(), PLearn::RBMConv2DConnection::computeProducts(), PLearn::computeRanks(), PLearn::LinearRegressor::computeResidualsVariance(), PLearn::LayerCostModule::computeSafeHisto(), PLearn::LocalizedFeaturesLayerVariable::computeSubsets(), PLearn::Kernel::computeTestGramMatrix(), PLearn::PruningLinearRegressor::computeTRatio(), PLearn::RBMModule::computeVisibleActivations(), PLearn::KLp0p1RBMModule::computeVisibleActivations(), PLearn::ICP::computeWeightedDistance(), PLearn::ChemicalICP::computeWeightedDistance(), PLearn::UniformizeLearner::computeWeightedRankMap(), PLearn::ICP::computeWeights(), PLearn::ChemicalICP::computeWeights(), PLearn::SumVarianceOfLinearTransformedBernoullis::computeWsqnorm(), PLearn::constrainedLinearRegression(), PLearn::convolve(), PLearn::convolve2D(), PLearn::convolve2Dbackprop(), PLearn::correlation(), PLearn::BallTreeNearestNeighbors::createAnchors(), PLearn::cross_valid(), PLearn::Variable::defineGradientLocation(), PLearn::Variable::defineValueLocation(), PLearn::SecondIterationWrapper::deGaussianize(), PLearn::det(), PLearn::diag(), PLearn::diagonalizedFactorsProduct(), PLearn::diagonalizedFactorsProductBprop(), PLearn::diagonalizedFactorsProductTranspose(), PLearn::diagonalizedFactorsProductTransposeBprop(), PLearn::diagonalizedFactorsTransposeProduct(), PLearn::diagonalizedFactorsTransposeProductBprop(), PLearn::diagonalizedFactorsTransposeProductTranspose(), PLearn::diagonalizedFactorsTransposeProductTransposeBprop(), PLearn::diagonalizeSubspace(), PLearn::diagonalOfSquare(), PLearn::diffSquareMultiplyAcc(), PLearn::GhostScript::displayBlack(), PLearn::GhostScript::displayGray(), PLearn::displayPoints(), PLearn::GhostScript::displayRGB(), PLearn::DX_write_2D_data(), PLearn::DX_write_2D_data_for_grid(), PLearn::ICP::dynamicDistanceThreshold(), PLearn::eigen_SymmMat(), PLearn::eigenSparseNonSymmMat(), PLearn::eigenSparseSymmMat(), PLearn::eigenVecOfSymmMat(), PLearn::DenoisingRecurrentNet::encode_artificialData(), PLearn::DenoisingRecurrentNet::encode_onehot_diffNote_duration(), PLearn::DenoisingRecurrentNet::encode_onehot_note_octav_duration(), PLearn::DenoisingRecurrentNet::encode_onehot_timeframe(), PLearn::DenoisingRecurrentNet::encodeAndCreateSupervisedSequence(), PLearn::DenoisingRecurrentNet::encodeAndCreateSupervisedSequence2(), PLearn::equals(), PLearn::Kernel::estimateHistograms(), PLearn::ReconstructionWeightsKernel::evaluate_sum_k_i_k_j(), PLearn::PLMPI::exchangeColumnBlocks(), PLearn::externalProduct(), PLearn::externalProductAcc(), PLearn::externalProductDivUpdate(), PLearn::externalProductMultUpdate(), PLearn::externalProductScaleAcc(), PLearn::RBMLateralBinomialLayer::externalSymetricProductAcc(), PLearn::StackedAutoassociatorsNet::fantasizeKTimeOnMultiSrcImg(), PLearn::DeepBeliefNet::fantasizeKTimeOnMultiSrcImg(), PLearn::GaussianProcessNLLVariable::fbpropFragments(), PLearn::NatGradSMPNNet::fbpropLoss(), PLearn::NatGradNNet::fbpropLoss(), PLearn::mNNet::fbpropLoss(), PLearn::Gnuplot::featureplot(), PLearn::PRandom::fill_random_normal(), PLearn::PRandom::fill_random_uniform(), PLearn::fillDiagonal(), PLearn::fillItSymmetric(), PLearn::NeuralProbabilisticLanguageModel::fillWeights(), PLearn::FeatureSetSequentialCRF::fillWeights(), PLearn::FeatureSetNNet::fillWeights(), PLearn::LiftStatsCollector::finalize(), PLearn::findSmallestEigenPairOfSymmMat(), PLearn::RBMLateralBinomialLayer::forget(), PLearn::PseudolikelihoodRBM::forget(), PLearn::GradNNetLayerModule::forget(), PLearn::WeightedLogGaussian::fprop(), PLearn::VBoundDBN2::fprop(), PLearn::TreeDBNModule::fprop(), PLearn::TransposedDoubleProductVariable::fprop(), PLearn::TraceVariable::fprop(), PLearn::TanhModule::fprop(), PLearn::SumVarianceOfLinearTransformedCategoricals::fprop(), PLearn::SumVarianceOfLinearTransformedBernoullis::fprop(), PLearn::SumEntropyOfCategoricals::fprop(), PLearn::SumEntropyOfBernoullis::fprop(), PLearn::SplitModule::fprop(), PLearn::SoftmaxNLLCostModule::fprop(), PLearn::SoftmaxModule::fprop(), PLearn::ShuntingNNetLayerModule::fprop(), PLearn::ScaleGradientModule::fprop(), PLearn::RunICPVariable::fprop(), PLearn::RBMWoodsLayer::fprop(), PLearn::RBMModule::fprop(), PLearn::RBMMixedLayer::fprop(), PLearn::RBMLayer::fprop(), PLearn::RBMLateralBinomialLayer::fprop(), PLearn::RBMConnection::fprop(), PLearn::RBMClassificationModule::fprop(), PLearn::RBMBinomialLayer::fprop(), PLearn::ProjectionErrorVariable::fprop(), PLearn::ProcessInputCostModule::fprop(), PLearn::OnBagsModule::fprop(), PLearn::NonDiagVariable::fprop(), PLearn::NllSemisphericalGaussianVariable::fprop(), PLearn::NllGeneralGaussianVariable::fprop(), PLearn::NLLCostModule::fprop(), PLearn::NetworkModule::fprop(), PLearn::MultiSampleVariable::fprop(), PLearn::MultiMaxVariable::fprop(), PLearn::MatrixModule::fprop(), PLearn::MatrixAffineTransformFeedbackVariable::fprop(), PLearn::LogSoftSoftMaxVariable::fprop(), PLearn::LocalizedFeaturesLayerVariable::fprop(), PLearn::LinearFilterModule::fprop(), PLearn::LayerCostModule::fprop(), PLearn::KLp0p1RBMModule::fprop(), PLearn::InsertZerosVariable::fprop(), PLearn::IdentityModule::fprop(), PLearn::GradNNetLayerModule::fprop(), PLearn::GaussianProcessNLLVariable::fprop(), PLearn::DoubleProductVariable::fprop(), PLearn::CrossEntropyCostModule::fprop(), PLearn::Cov2CorrVariable::fprop(), PLearn::CostModule::fprop(), PLearn::ConstrainVariable::fprop(), PLearn::ConcatColumnsVariable::fprop(), PLearn::CombiningCostsModule::fprop(), PLearn::ClassErrorCostModule::fprop(), PLearn::BinarizeModule::fprop(), PLearn::RBMWoodsLayer::fpropNLL(), PLearn::RBMMultinomialLayer::fpropNLL(), PLearn::RBMMixedLayer::fpropNLL(), PLearn::RBMLocalMultinomialLayer::fpropNLL(), PLearn::RBMLayer::fpropNLL(), PLearn::RBMLateralBinomialLayer::fpropNLL(), PLearn::RBMGaussianLayer::fpropNLL(), PLearn::RBMBinomialLayer::fpropNLL(), PLearn::SumOverBagsVariable::fpropOneBag(), PLearn::TreeDBNModule::full_fprop(), PLearn::GCV(), PLearn::GDFindSmallEigenPairs(), PLearn::generalizedEigenVecOfSymmMat(), PLearn::RBMDistribution::generateN(), PLearn::PDistribution::generateN(), PLearn::ExplicitListOracle::generateNextTrial(), PLearn::RBMWoodsLayer::generateSamples(), PLearn::RBMTruncExpLayer::generateSamples(), PLearn::RBMRateLayer::generateSamples(), PLearn::RBMMultinomialLayer::generateSamples(), PLearn::RBMLocalMultinomialLayer::generateSamples(), PLearn::RBMLateralBinomialLayer::generateSamples(), PLearn::RBMGaussianLayer::generateSamples(), PLearn::RBMBinomialLayer::generateSamples(), PLearn::geometric_mean(), PLearn::RegressionTreeRegisters::getAllRegisteredRow(), PLearn::RegressionTreeRegisters::getAllRegisteredRowLeave(), PLearn::VecStatsCollector::getCovariance(), PLearn::LemmatizeVMatrix::getLemma(), PLearn::VMatrixFromDistribution::getMat(), PLearn::VMatrix::getMat(), PLearn::TransposeVMatrix::getMat(), PLearn::SubVMatrix::getMat(), PLearn::MemoryVMatrix::getMat(), PLearn::RemapLastColumnVMatrix::getNewRow(), PLearn::PairsVMatrix::getNewRow(), PLearn::OnlineLearningModule::getPortSizes(), PLearn::MatrixModule::getPortSizes(), PLearn::CostModule::getPortSizes(), PLearn::ToBagSplitter::getSplit(), PLearn::BootstrapSplitter::getSplit(), PLearn::OnlineGramNaturalGradientOptimizer::gramEigenNaturalGradient(), PLearn::GramSchmidtOrthogonalization(), PLearn::grep(), PLearn::InferenceRBM::hiddenExpGivenInput(), PLearn::InferenceRBM::hiddenExpGivenInputTarget(), PLearn::identityMatrix(), PLearn::inverse(), PLearn::GaussianProcessRegressor::inverseCovTimesVec(), PLearn::InversePowerIteration(), PLearn::invertElements(), PLearn::TMat< T >::isEqual(), PLearn::ICP::iterate(), PLearn::ICP::iterativeReweight(), PLearn::jacobi(), PLearn::kernelPCAfromDotProducts(), PLearn::lapackCholeskyDecompositionInPlace(), PLearn::lapackCholeskySolveInPlace(), PLearn::lapackEIGEN(), PLearn::lapackGeneralizedEIGEN(), PLearn::lapackSolveLinearSystem(), PLearn::lapackSVD(), PLearn::layerBpropUpdate(), PLearn::layerL1BpropUpdate(), PLearn::layerL2BpropUpdate(), PLearn::leftPseudoInverse(), PLearn::ObservationWindow::length(), PLearn::StatsCollector::lift(), PLearn::linearRegression(), PLearn::linearRegressionNoBias(), PLearn::loadAsciiWithoutSize(), PLearn::loadATT800(), PLearn::loadBreastCancer(), PLearn::loadCorel(), PLearn::loadCorelDatamat(), PLearn::loadDiabetes(), PLearn::loadLetters(), PLearn::loadPMat(), PLearn::loadUCISet(), PLearn::loadUSPS(), PLearn::ManifoldKNNDistribution::log_density(), PLearn::log_fullGaussianRBF(), PLearn::LU_decomposition(), main(), PLearn::makeItSymmetric(), PLearn::makeRowsSumTo1(), PLearn::ChemicalICP::matchNearestNeighbors(), PLearn::matColumnDotVec(), PLearn::matInvert(), PLearn::matlabR11eigs(), PLearn::matlabSave(), PLearn::matRowsDots(), PLearn::matRowsDotsAcc(), PLearn::max(), PLearn::maxabs(), PLearn::maxPointMotion(), PLearn::mean(), PLearn::StatsCollector::mean_lift(), PLearn::VecStatsCollector::merge(), PLearn::metricMultiDimensionalScaling(), PLearn::min(), PLearn::minabs(), PLearn::GaussMix::missingExpectation(), PLearn::Molecule::Molecule(), PLearn::multiply(), PLearn::multiplyAcc(), PLearn::multiplyColumns(), PLearn::multiplyScaledAdd(), PLearn::OnlineLearningModule::namedBpropAccUpdate(), PLearn::PLearnDiff::nDiffs(), PLearn::negateElements(), PLearn::PruningLinearRegressor::newDatasetIndices(), PLearn::newIndexedMatArray(), PLearn::ConvertToPyObject< Mat >::newPyObject(), PLearn::RBMTrainer::NLL(), PLearn::RGBImagesVMatrix::nObjects(), PLearn::normalize(), PLearn::normalizeRows(), PLearn::FractionSplitter::nsplits(), PLearn::SubsamplingDBN::onlineStep(), PLearn::StackedAutoassociatorsNet::onlineStep(), PLearn::NatGradSMPNNet::onlineStep(), PLearn::NatGradNNet::onlineStep(), PLearn::mNNet::onlineStep(), PLearn::DeepBeliefNet::onlineStep(), PLearn::CorrelationProfiler::operator()(), PLearn::operator*(), PLearn::operator*=(), PLearn::operator+(), PLearn::operator+=(), PLearn::operator-(), PLearn::operator-=(), PLearn::operator/=(), PLearn::operator<<(), PLearn::TMat< T >::operator==(), PLearn::operator^(), PLearn::GradientOptimizer::optimizeN(), PLearn::AutoScaledGradientOptimizer::optimizeN(), PLearn::partialSortRows(), PLearn::RBMMatrixConnection::petiteCulotteOlivierCD(), PLearn::Gnuplot::plotClasses(), PLearn::YMDDatedVMatrix::positionOfDate(), PLearn::PowerIteration(), PLearn::product(), PLearn::product2Acc(), PLearn::product2Transpose(), PLearn::product2TransposeAcc(), PLearn::productAcc(), PLearn::productScaleAcc(), PLearn::productTranspose(), PLearn::productTransposeAcc(), PLearn::productTransposeScaleAcc(), PLearn::projectOnOrthogonalSubspace(), PLearn::VMatrix::putMat(), PLearn::SubVMatrix::putMat(), PLearn::MemoryVMatrix::putMat(), PLearn::GaussianProcessRegressor::QFormInverse(), PLearn::qld_interface(), PLearn::SourceVariable::randomInitialize(), PLearn::Molecule::readMolecule(), PLearn::RBMTrainer::recError(), PLearn::DenoisingRecurrentNet::recurrentUpdate(), PLearn::regulargrid_x_y_outputs_to_bitmap(), PLearn::regulargrid_x_y_rgbreal_to_bitmap(), PLearn::regularizeMatrix(), PLearn::PLearner::remote_computeOutputs(), PLearn::remote_solveLinearSystemByCholesky(), PLearn::removeRow(), PLearn::RGBImagesVMatrix::RGBImagesVMatrix(), PLearn::ridgeRegressionByGCV(), PLearn::rightPseudoInverse(), PLearn::rotationFromWeightedMatchedPoints(), PLearn::rowArgmax(), PLearn::rowArgmin(), PLearn::rowMax(), PLearn::rowMean(), PLearn::rowMin(), PLearn::rowSum(), PLearn::rowSumAcc(), PLearn::rowSumOfSquares(), PLearn::rowVariance(), PLearn::ExtractOptionCommand::run(), PLearn::ChemicalICP::run(), PLearn::RGBImagesVMatrix::sample(), PLearn::sample(), PLearn::saveAscii(), PLearn::saveAsciiWithoutSize(), PLearn::saveGnuplot(), PLearn::savePMat(), PLearn::saveSNMat(), PLearn::scores_to_winners(), PLearn::RGBImagesVMatrix::seek(), PLearn::select(), PLearn::selectAndOrder(), PLearn::selectRows(), PLearn::ThresholdedKernel::setDataForKernelMatrix(), PLearn::MemoryCachedKernel::setDataForKernelMatrix(), PLearn::RBMMixedLayer::setExpectations(), PLearn::RBMLayer::setExpectations(), PLearn::RBMMixedLayer::setExpectationsByRef(), PLearn::RBMLayer::setExpectationsByRef(), PLearn::MatrixModule::setGradientTo(), PLearn::TransformationLearner::setTransformsParameters(), PLearn::Variable::setValueSubMat(), PLearn::SurfaceMesh::setVertexCoords(), PLearn::SurfaceMesh::setVertexCoordsAndFeatures(), PLearn::SurfaceMesh::setVertexFeatures(), PLearn::SurfaceMesh::setVertexNorms(), PLearn::shuffleRows(), PLearn::PRandom::shuffleRows(), PLearn::smooth(), PLearn::smoothCorelHisto(), PLearn::softmax(), PLearn::solveLinearSystem(), PLearn::solveLinearSystemByCholesky(), PLearn::solveTransposeLinearSystemByCholesky(), PLearn::sortColumns(), PLearn::sortRows(), PLearn::RegressionTreeRegisters::sortRows(), PLearn::SpearmanRankCorrelation(), PLearn::DenoisingRecurrentNet::splitRawMaskedSupervisedSequence(), PLearn::sqrt(), PLearn::square(), PLearn::squareMultiplyAcc(), PLearn::squareProductAcc(), PLearn::squareProductTranspose(), PLearn::squareProductTransposeAcc(), PLearn::subsample(), PLearn::substract(), PLearn::substractFromColumns(), PLearn::substractFromRows(), PLearn::sum(), PLearn::sum_of_squares(), PLearn::InferenceRBM::supCDStep(), PLearn::InferenceRBM::targetExpGivenInput(), PLearn::testSpearmanRankCorrelation(), PLearn::ThresholdedKernel::thresholdGramMatrix(), PLearn::TMat< T >::TMat(), PLearn::SparseMatrix::toMat(), PLearn::trace(), PLearn::StructuralLearner::train(), PLearn::RankLearner::train(), PLearn::PseudolikelihoodRBM::train(), PLearn::ModuleLearner::train(), PLearn::LLC::train(), PLearn::LinearRegressor::train(), PLearn::KernelRidgeRegressor::train(), PLearn::KernelProjection::train(), PLearn::GaussMix::train(), PLearn::GaussianProcessRegressor::train(), PLearn::GaussianDistribution::train(), PLearn::DiverseComponentAnalysis::train(), PLearn::ConditionalDensityNet::train(), PLearn::AutoLinearRegressor::train(), PLearn::ModuleLearner::trainingStep(), PLearn::GaussianProcessRegressor::trainProjectedProcess(), PLearn::DenoisingRecurrentNet::trainUnconditionalPredictor(), PLearn::transformPoints(), PLearn::transpose(), PLearn::transposedLayerL1BpropUpdate(), PLearn::transposedLayerL2BpropUpdate(), PLearn::transposeProduct(), PLearn::transposeProduct2(), PLearn::transposeProduct2Acc(), PLearn::transposeProductAcc(), PLearn::transposeProductScaleAcc(), PLearn::transposeTransposeProduct(), PLearn::transposeTransposeProductAcc(), PLearn::transposeTransposeProductScaleAcc(), PLearn::VecStatsCollector::update(), PLearn::StatsIterator::update(), PLearn::StatsItArray::update(), PLearn::LiftStatsIterator::update(), PLearn::RBMSparse1DMatrixConnection::update(), PLearn::RBMMatrixTransposeConnection::update(), PLearn::RBMMatrixConnectionNatGrad::update(), PLearn::RBMMatrixConnection::update(), PLearn::RBMLLParameters::update(), PLearn::RBMLayer::update(), PLearn::RBMLateralBinomialLayer::update(), PLearn::RBMGaussianLayer::update(), PLearn::RBMDiagonalMatrixConnection::update(), PLearn::RBMConv2DConnection::update(), PLearn::LiftStatsCollector::update(), PLearn::NeuralProbabilisticLanguageModel::update_affine_transform(), PLearn::FeatureSetSequentialCRF::update_affine_transform(), PLearn::FeatureSetNNet::update_affine_transform(), PLearn::RBMMatrixConnection::updateCDandGibbs(), PLearn::RBMLayer::updateCDandGibbs(), PLearn::GaussMix::updateCholeskyFromPrevious(), PLearn::RBMMatrixConnection::updateGibbs(), PLearn::RBMLayer::updateGibbs(), PLearn::GaussMix::updateInverseVarianceFromPrevious(), PLearn::PLearner::use(), PLearn::Learner::use(), PLearn::variance(), PLearn::GaussianProcessRegressor::variance(), PLearn::NeuralProbabilisticLanguageModel::verify_gradient_affine_transform(), PLearn::FeatureSetSequentialCRF::verify_gradient_affine_transform(), PLearn::FeatureSetNNet::verify_gradient_affine_transform(), PLearn::weightedCentroid(), PLearn::weightedLinearRegression(), PLearn::weightedRidgeRegressionByGCV(), PLearn::weightedRotationFromMatchedPoints(), PLearn::weightedTransformationFromMatchedPoints(), PLearn::GhostScript::writeBitmapHexString1Bit(), PLearn::GhostScript::writeBitmapHexString24Bits(), and PLearn::GhostScript::writeBitmapHexString8Bits().

    { return length_; }
template<class T >
void PLearn::TMat< T >::makeDeepCopyFromShallowCopy ( CopiesMap copies)

************ Deep copying

Definition at line 464 of file TMat_impl.h.

References PLearn::deepCopyField().

Referenced by PLearn::TMat< T >::deepCopy(), and PLearn::deepCopyField().

{
    deepCopyField(storage, copies);
}

Here is the call graph for this function:

Here is the caller graph for this function:

template<class T>
void PLearn::TMat< T >::makeSharedValue ( T *  x,
int  n 
) [inline]

make the storage point to this address and copy current value to it (i.e. without changing current contents)

< get data start

Definition at line 841 of file TMat_decl.h.

Referenced by PLearn::RBMMatrixConnection::makeParametersPointHere(), PLearn::RBMQLParameters::makeParametersPointHere(), PLearn::RBMConv2DConnection::makeParametersPointHere(), PLearn::RBMLLParameters::makeParametersPointHere(), PLearn::RBMConv2DLLParameters::makeParametersPointHere(), PLearn::RBMGenericParameters::makeParametersPointHere(), and PLearn::RBMLQParameters::makeParametersPointHere().

    {
#ifdef BOUNDCHECK
        int m = size();
        if(n != m)
            PLERROR("IN TMat::makeSharedValue(T* x, int n)\nn(%d)!=size(%d)",
                    n,m);
        if(offset_!=0)
            PLERROR("IN TMat::makeSharedValue(T* x, int n)\noffset should be 0.");
        if (mod_!=width_)
            PLERROR("IN TMat::makeSharedValue(T* x, int n)\nMatrix should be compact (mod==width), but isn't.");
#endif
        T* v = data(); 
        for(int i=0,k=0; i<length_; i++, v+=mod_)
            for (int j=0;j<width_; j++, k++)
                x[k] = v[j];
        storage->pointTo(n,x);
    }

Here is the caller graph for this function:

template<class T>
int PLearn::TMat< T >::mod ( ) const [inline]

Definition at line 254 of file TMat_decl.h.

Referenced by PLearn::absargmax(), PLearn::add(), PLearn::addToColumns(), PLearn::addToDiagonal(), PLearn::argmax(), PLearn::argmin(), PLearn::averageAcrossRowsAndColumns(), PLearn::backConvolve2D(), PLearn::backConvolve2Dbackprop(), PLearn::SparseIncrementalAffineTransformVariable::bprop(), PLearn::FNetLayerVariable::bprop(), PLearn::ConcatColumnsVariable::bprop(), PLearn::CostModule::bpropAccUpdate(), PLearn::RBMLateralBinomialLayer::bpropNLL(), PLearn::RBMLateralBinomialLayer::bpropUpdate(), PLearn::chol_rotapp_tr_opt(), PLearn::choleskyDecomposition(), PLearn::choleskySolve(), PLearn::SquaredExponentialARDKernel::computeGramMatrix(), PLearn::RationalQuadraticARDKernel::computeGramMatrix(), PLearn::NeuralNetworkARDKernel::computeGramMatrix(), PLearn::Matern1ARDKernel::computeGramMatrix(), PLearn::LinearARDKernel::computeGramMatrix(), PLearn::KroneckerBaseKernel::computeGramMatrix(), PLearn::Kernel::computeGramMatrix(), PLearn::IIDNoiseKernel::computeGramMatrix(), PLearn::RationalQuadraticARDKernel::computeGramMatrixDerivIspAlpha(), PLearn::SquaredExponentialARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::RationalQuadraticARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::NeuralNetworkARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::Matern1ARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::LinearARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::MemoryCachedKernel::computeGramMatrixNV(), PLearn::computeInverseStandardDeviationFromMeanAndSquareMean(), PLearn::GaussianProcessRegressor::computeOutputCovMat(), PLearn::Kernel::computePartialGramMatrix(), PLearn::convolve(), PLearn::convolve2D(), PLearn::convolve2Dbackprop(), PLearn::det(), PLearn::eigen_SymmMat(), PLearn::fillItSymmetric(), PLearn::RBMLateralBinomialLayer::forget(), PLearn::FNetLayerVariable::fprop(), PLearn::ConcatColumnsVariable::fprop(), PLearn::geometric_mean(), PLearn::invertElements(), PLearn::lapackCholeskyDecompositionInPlace(), PLearn::lapackCholeskySolveInPlace(), PLearn::lapackEIGEN(), PLearn::lapackGeneralizedEIGEN(), PLearn::lapackSolveLinearSystem(), PLearn::lapackSVD(), PLearn::linearRegression(), PLearn::LU_decomposition(), PLearn::matColumnDotVec(), PLearn::matInvert(), PLearn::max(), PLearn::maxabs(), PLearn::mean(), PLearn::min(), PLearn::minabs(), PLearn::multiplyScaledAdd(), PLearn::negateElements(), PLearn::ConvertToPyObject< Mat >::newPyObject(), PLearn::operator*=(), PLearn::operator+=(), PLearn::operator-(), PLearn::operator-=(), PLearn::operator/=(), PLearn::partialSortRows(), PLearn::RBMMatrixConnection::petiteCulotteOlivierCD(), PLearn::product(), PLearn::product2Acc(), PLearn::productAcc(), PLearn::productScaleAcc(), PLearn::regularizeMatrix(), PLearn::sortRows(), PLearn::SparseMatrix::SparseMatrix(), PLearn::squareProductAcc(), PLearn::subsample(), PLearn::substract(), PLearn::substractFromColumns(), PLearn::sum(), PLearn::sum_of_squares(), PLearn::PseudolikelihoodRBM::train(), PLearn::GaussianProcessRegressor::train(), PLearn::transpose(), PLearn::transposeProduct(), PLearn::transposeProduct2(), PLearn::transposeProduct2Acc(), PLearn::transposeProductAcc(), PLearn::transposeProductScaleAcc(), PLearn::transposeTransposeProduct(), PLearn::transposeTransposeProductAcc(), PLearn::transposeTransposeProductScaleAcc(), PLearn::RBMMatrixTransposeConnection::update(), PLearn::RBMMatrixConnection::update(), PLearn::RBMLLParameters::update(), PLearn::RBMLateralBinomialLayer::update(), PLearn::RBMConv2DLLParameters::update(), PLearn::RBMConv2DConnection::update(), and PLearn::variance().

    { return mod_; }
template<class T>
int PLearn::TMat< T >::ncols ( ) const [inline]

Definition at line 81 of file TMat_decl.h.

Referenced by calcTransformation4(), findRelevantWeights(), nodekernel(), and performLP().

{ return width_; }

Here is the caller graph for this function:

template<class T>
int PLearn::TMat< T >::nrows ( ) const [inline]

for template compatibility with other types of matrices

Definition at line 80 of file TMat_decl.h.

Referenced by autoThreshLP(), calcTransformation4(), calculateEuclDist(), findRelevantWeights(), nodekernel(), and performLP().

{ return length_; }

Here is the caller graph for this function:

template<class T>
bool PLearn::TMat< T >::operator! ( ) const [inline]

To allow if(!m) statements.

This method is commented out because it is much too dangerous: it allows a TMat to be converted into an int, which can cause some very weird bugs that the compiler would have caught otherwise.

Definition at line 747 of file TMat_decl.h.

    { return isEmpty(); }
template<class T>
T& PLearn::TMat< T >::operator() ( int  rownum,
int  colnum 
) const [inline]

Definition at line 345 of file TMat_decl.h.

    {
#ifdef BOUNDCHECK
        if(rownum<0 || rownum>=length() || colnum<0 || colnum>=width())
            PLERROR("OUT OF BOUND ACCESS IN TMat::operator()(int rownum, int colnum)"
                    " width=%d; length=%d; colnum=%d; rownum=%d;", width(), length(), colnum, rownum);
#endif
        return storage->data[offset_ + mod()*rownum + colnum];
    }
template<class T>
template<class I >
TMat<T> PLearn::TMat< T >::operator() ( const TVec< I > &  rows,
const TVec< I > &  columns 
) const [inline]

Definition at line 625 of file TMat_decl.h.

    {
        TMat<T> result(rows.length(),columns.length());
        select(*this,rows,columns,result);
        return result;
    }
template<class T>
TVec<T> PLearn::TMat< T >::operator() ( int  rownum) const [inline]

Definition at line 355 of file TMat_decl.h.

    {
#ifdef BOUNDCHECK
        if(rownum<0 || rownum>=length())
            PLERROR("OUT OF BOUND ACCESS IN TMat_impl::operator()(int rownum)");
#endif
        TVec<T> tv;
        tv.length_ = width();
        tv.offset_ = offset_ + mod()*rownum;
        tv.storage = storage;
        return tv;
    }
template<class T>
void PLearn::TMat< T >::operator<< ( const string &  datastring) const [inline]

Definition at line 949 of file TMat_decl.h.

    { 
        // istrstream in(datastring.c_str());
        PStream in = openString(datastring,PStream::plearn_ascii);
        input(in); 
    }
template<class T>
void PLearn::TMat< T >::operator= ( const T &  f) const [inline]

Definition at line 768 of file TMat_decl.h.

    { fill(f); }
template<class T>
const TMat<T>& PLearn::TMat< T >::operator= ( const TMat< T > &  other) [inline]

NOTE: operator= COPIES THE TMat STRUCTURE BUT NOT THE DATA (use operator<< to copy data)

Definition at line 119 of file TMat_decl.h.

    {
        storage = other.storage;
        offset_ = other.offset_;
        mod_ = other.mod_;
        length_ = other.length_;
        width_ = other.width_;
        return *this;
    }
template<class T>
bool PLearn::TMat< T >::operator== ( const TMat< T > &  other) const [inline]

Definition at line 523 of file TMat_impl.h.

References PLearn::TMat< T >::begin(), PLearn::TMat< T >::length(), and PLearn::TMat< T >::width().

{
    if ( length() != other.length() || width() != other.width() )
        return false;
  
    iterator it       = begin();
    iterator end_     = end();
    iterator other_it = other.begin();

    for(; it != end_; ++it, ++other_it)
        if(*it != *other_it)
            return false;
  
    return true;
}

Here is the call graph for this function:

template<class T>
T* PLearn::TMat< T >::operator[] ( int  rownum) const [inline]

Returns a pointer to the data beginning of the required row.

Definition at line 334 of file TMat_decl.h.

    {
#ifdef BOUNDCHECK
        if(rownum<0 || rownum>=length())
            PLERROR("OUT OF BOUND ACCESS IN TMat::operator[](int rownum=%d), length=%d",rownum,length());
#endif
        return storage->data + offset_ + mod()*rownum; 
    }
template<class T>
void PLearn::TMat< T >::pop_back ( ) [inline]

Definition at line 834 of file TMat_decl.h.

{ length_ -= 1; }
template<class T >
void PLearn::TMat< T >::print ( ostream &  out = cout) const

C++ stream output.

Definition at line 342 of file TMat_impl.h.

References i, j, and PLearn::left().

Referenced by PLearn::CorrelationProfiler::printAndReset().

{
    out.flags(ios::left);
    for(int i=0; i<length(); i++)
    {
        const T* m_i = rowdata(i);
        for(int j=0; j<width(); j++)
            out << setw(11) << m_i[j] << ' ';
        out << "\n";
    }
    out.flush();
}

Here is the call graph for this function:

Here is the caller graph for this function:

template<class T>
void PLearn::TMat< T >::push_back ( const TVec< T > &  newrow) [inline]

stl-like push_back and pop_back

Definition at line 833 of file TMat_decl.h.

{ appendRow(newrow); }
template<class T>
void PLearn::TMat< T >::read ( PStream in) [inline]

reads the Mat from the PStream: Note that users should rather use the form in >> m;

Definition at line 470 of file TMat_decl.h.

Referenced by PLearn::operator>>().

    {

        switch(in.inmode)
        {
        case PStream::raw_ascii:
        case PStream::raw_binary:
        {
            T* ptr = (length_>0 && width_>0)? data():0;
            for(int i=0; i<length_; i++, ptr+=mod_)
                for(int j=0; j<width_; j++)
                    in >> ptr[j];
        }
        break;

        case PStream::plearn_ascii:
        case PStream::plearn_binary:
        {
            in.skipBlanksAndComments();
            int c = in.peek();
            if(c=='T') // explicit storage
            {
                char word[6];
                // !!!! BUG: For some reason, this hangs!!!
                // in.read(word,5);

                for(int i=0; i<5; i++)
                    in.get(word[i]);

                word[5]='\0';
                if(strcmp(word,"TMat(")!=0)
                    PLERROR("In operator>>(PStream&, TMat&) '%s' not a proper header for a TMat!",word);
                // v.storage = 0;
                in >> length_ >> width_ >> mod_ >> offset_;
                in >> storage;
                in.skipBlanksAndCommentsAndSeparators();
                int c = in.get(); // skip ')'
                if(c!=')')
                    PLERROR("In operator>>(PStream&, TMat&) expected a closing parenthesis, found '%c'",c);
            }
            else // implicit storage
            {
                if(isdigit(c)) // ascii mode with length and width given  
                {
                    int l,w;
                    in >> l >> w;
                    in.skipBlanksAndComments();
                    c = in.get();
                    if(c!='[')
                        PLERROR("Error in TMat::read(PStream& in), expected '[', read '%c'",c);
                    in.skipBlanksAndCommentsAndSeparators();
                    resize(l,w);
                    T* ptr = (l>0 && w>0)? data():0;
                    for(int i=0; i<length_; i++, ptr+=mod_)
                        for(int j=0; j<width_; j++)
                        {
                            in.skipBlanksAndCommentsAndSeparators();
                            try{
                                in >> ptr[j];
                            }
                            catch(const PLearnError& e) {
                                PLERROR("In TMat::read() - Error while reading a serialised TMat<T>."
                                        " Did you set correctly the mat size?\n"
                                        "%s",e.message().c_str());
                            }
                        }
                    in.skipBlanksAndCommentsAndSeparators();
                    c = in.get();
                    if(c!=']')
                        PLERROR("Error in TMat::read(PStream& in), expected ']', read '%c'",c);
                }
                else if(c==0x14 || c==0x15) // it's a binary 2D sequence
                {
                    in.get(); // eat c
                    unsigned char typecode = in.get(); 
                    int l, w;                  
                    in.read((char*)&l,sizeof(l));
                    in.read((char*)&w,sizeof(w));
                    bool inverted_byte_order = ((c==0x14 && byte_order()==BIG_ENDIAN_ORDER) 
                                                || (c==0x15 && byte_order()==LITTLE_ENDIAN_ORDER) );
                    if(inverted_byte_order)
                    {
                        endianswap(&l);
                        endianswap(&w);
                    }
                    resize(l,w);
                    T* ptr = (l>0 && w>0)? data():0;
                    for(int i=0; i<length_; i++, ptr+=mod_)                    
                        binread_(in, ptr, width_, typecode);
                }
                else
                    PLERROR("In TMat::read(PStream& in) Char with ascii code %d not a proper first character in the header of a TMat!",c);
            }
        }
        break;
      
        default:
            PLERROR("In TMat<T>::read(PStream& in)  unknown inmode!!!!!!!!!");
            break;
        }
    }

Here is the caller graph for this function:

template<class T>
void PLearn::TMat< T >::resize ( int  new_length,
int  new_width,
int  extra = 0,
bool  preserve_content = false 
) [inline]

Resizes the matrix to a new length() and width().

Note that the previous structure of the data in the matrix is not preserved if you increase the width() beyond mod(). The underlying storage is never shrunk, and it is grown only if necessary. When grown, it is grown with extra entries to anticipate further resizes. If preserve_content is true then a change of mod_ triggers a COPY of the old entries so that their old value remains accessed at the same indices.

This function is split into several parts: a `small' one that handles the common cases, and a few `auxiliary' ones that perform the heavy-lifting. The small one is easily inlined, whereas having one `large' resize() function would overflow the compiler inlining threshold, yielding no inlining whatsoever.

Definition at line 200 of file TMat_decl.h.

Referenced by PLearn::addEigenMatrices(), PLearn::GaussMix::addToCovariance(), PLearn::GaussianProcessNLLVariable::alpha(), PLearn::BallTreeNearestNeighbors::anchorTrain(), PLearn::BasisSelectionRegressor::appendFunctionToSelection(), PLearn::BasisSelectionRegressor::appendKernelFunctions(), PLearn::Kernel::apply(), PLearn::applyGeomTransformation(), PLearn::autocorrelation_function(), PLearn::bootstrap_rows(), PLearn::LogaddOnBagsModule::bprop(), PLearn::VBoundDBN2::bpropAccUpdate(), PLearn::SplitModule::bpropAccUpdate(), PLearn::SoftmaxNLLCostModule::bpropAccUpdate(), PLearn::RBMSparse1DMatrixConnection::bpropAccUpdate(), PLearn::RBMModule::bpropAccUpdate(), PLearn::RBMMixedConnection::bpropAccUpdate(), PLearn::RBMMatrixConnection::bpropAccUpdate(), PLearn::RBMConv2DConnection::bpropAccUpdate(), PLearn::OnlineLearningModule::bpropAccUpdate(), PLearn::OnBagsModule::bpropAccUpdate(), PLearn::NLLCostModule::bpropAccUpdate(), PLearn::NetworkModule::bpropAccUpdate(), PLearn::MaxSubsampling2DModule::bpropAccUpdate(), PLearn::MatrixModule::bpropAccUpdate(), PLearn::KLp0p1RBMModule::bpropAccUpdate(), PLearn::CrossEntropyCostModule::bpropAccUpdate(), PLearn::CostModule::bpropAccUpdate(), PLearn::Convolution2DModule::bpropAccUpdate(), PLearn::BinarizeModule::bpropAccUpdate(), PLearn::bpropCholeskyDecomposition(), PLearn::RBMWoodsLayer::bpropNLL(), PLearn::RBMMultinomialLayer::bpropNLL(), PLearn::RBMMixedLayer::bpropNLL(), PLearn::RBMLocalMultinomialLayer::bpropNLL(), PLearn::RBMLateralBinomialLayer::bpropNLL(), PLearn::RBMGaussianLayer::bpropNLL(), PLearn::RBMBinomialLayer::bpropNLL(), PLearn::TanhModule::bpropUpdate(), PLearn::SquaredErrorCostModule::bpropUpdate(), PLearn::SoftmaxNLLCostModule::bpropUpdate(), PLearn::SoftmaxModule::bpropUpdate(), PLearn::ShuntingNNetLayerModule::bpropUpdate(), PLearn::ScaleGradientModule::bpropUpdate(), PLearn::RBMWoodsLayer::bpropUpdate(), PLearn::RBMSparse1DMatrixConnection::bpropUpdate(), PLearn::RBMMultinomialLayer::bpropUpdate(), PLearn::RBMMixedLayer::bpropUpdate(), PLearn::RBMMixedConnection::bpropUpdate(), PLearn::RBMMatrixTransposeConnection::bpropUpdate(), PLearn::RBMMatrixConnectionNatGrad::bpropUpdate(), PLearn::RBMMatrixConnection::bpropUpdate(), PLearn::RBMLocalMultinomialLayer::bpropUpdate(), PLearn::RBMLateralBinomialLayer::bpropUpdate(), PLearn::RBMDiagonalMatrixConnection::bpropUpdate(), PLearn::RBMConv2DConnection::bpropUpdate(), PLearn::RBMBinomialLayer::bpropUpdate(), PLearn::OnlineLearningModule::bpropUpdate(), PLearn::OnBagsModule::bpropUpdate(), PLearn::NLLCostModule::bpropUpdate(), PLearn::ModuleStackModule::bpropUpdate(), PLearn::LinearFilterModule::bpropUpdate(), PLearn::LayerCostModule::bpropUpdate(), PLearn::GradNNetLayerModule::bpropUpdate(), PLearn::CrossEntropyCostModule::bpropUpdate(), PLearn::CombiningCostsModule::bpropUpdate(), PLearn::VMatrixFromDistribution::build_(), PLearn::TrainValidTestSplitter::build_(), PLearn::Supersampling2DModule::build_(), PLearn::SumOverBagsVariable::build_(), PLearn::Subsampling2DModule::build_(), PLearn::StackedSVDNet::build_(), PLearn::SplitWiseValidationVMatrix::build_(), PLearn::SparseIncrementalAffineTransformVariable::build_(), PLearn::SequentialModelSelector::build_(), PLearn::SequentialLearner::build_(), PLearn::RemoveDuplicateVMatrix::build_(), PLearn::RBMSparse1DMatrixConnection::build_(), PLearn::RBMQLParameters::build_(), PLearn::RBMMultitaskClassificationModule::build_(), PLearn::RBMModule::build_(), PLearn::RBMMatrixTransposeConnection::build_(), PLearn::RBMMatrixConnectionNatGrad::build_(), PLearn::RBMMatrixConnection::build_(), PLearn::RBMLQParameters::build_(), PLearn::RBMLLParameters::build_(), PLearn::RBMLateralBinomialLayer::build_(), PLearn::RBMGenericParameters::build_(), PLearn::RBMConv2DLLParameters::build_(), PLearn::RBMConv2DConnection::build_(), PLearn::RBMConnection::build_(), PLearn::RankedVMatrix::build_(), PLearn::RandomGaussMix::build_(), PLearn::ProjectionErrorVariable::build_(), PLearn::PLearnDiff::build_(), PLearn::OnlineGramNaturalGradientOptimizer::build_(), PLearn::OnBagsModule::build_(), PLearn::NxProfileLearner::build_(), PLearn::NonLocalManifoldParzen::build_(), PLearn::NNet::build_(), PLearn::NllSemisphericalGaussianVariable::build_(), PLearn::NllGeneralGaussianVariable::build_(), PLearn::NetworkModule::build_(), PLearn::NeighborhoodImputationVMatrix::build_(), PLearn::NatGradSMPNNet::build_(), PLearn::NatGradNNet::build_(), PLearn::MultiTargetOneHotVMatrix::build_(), PLearn::MultiInstanceVMatrix::build_(), PLearn::MovingAverageVMatrix::build_(), PLearn::ModuleTester::build_(), PLearn::mNNet::build_(), PLearn::MemoryVMatrix::build_(), PLearn::MaxSubsampling2DModule::build_(), PLearn::ManifoldKNNDistribution::build_(), PLearn::LocalNeighborsDifferencesVMatrix::build_(), PLearn::LayerCostModule::build_(), PLearn::KNNVMatrix::build_(), PLearn::KNNImputationVMatrix::build_(), PLearn::KLp0p1RBMModule::build_(), PLearn::IncrementalNNet::build_(), PLearn::GraphicalBiText::build_(), PLearn::GramVMatrix::build_(), PLearn::GaussianProcessRegressor::build_(), PLearn::GaussianContinuumDistribution::build_(), PLearn::GaussianContinuum::build_(), PLearn::FNetLayerVariable::build_(), PLearn::ExtractNNetParamsVMatrix::build_(), PLearn::EntropyContrast::build_(), PLearn::DictionaryVMatrix::build_(), PLearn::CumVMatrix::build_(), PLearn::CorrelationProfiler::build_(), PLearn::Convolution2DModule::build_(), PLearn::CompareLearner::build_(), PLearn::BootstrapSplitter::build_(), PLearn::BaseRegressorConfidence::build_(), PLearn::BackConvolution2DModule::build_(), PLearn::AppendNeighborsVMatrix::build_(), PLearn::AddCostToLearner::build_(), PLearn::Convolution2DModule::build_kernels(), PLearn::BackConvolution2DModule::build_kernels(), PLearn::PseudolikelihoodRBM::build_layers_and_connections(), PLearn::RBMJointLLParameters::build_units_types(), PLearn::RBMJointGenericParameters::build_units_types(), PLearn::ChemicalICP::cacheFeatureDistances(), PLearn::RBMTrainer::CD1(), PLearn::chol_rotapp_tr(), PLearn::chol_rotapp_tr_opt(), PLearn::choleskyAppendDimension(), PLearn::choleskyDecomposition(), PLearn::choleskyInsertBasis(), PLearn::choleskyInvert(), PLearn::LayerCostModule::computeCorrelationStatistics(), PLearn::AddCostToLearner::computeCostsFromOutputs(), PLearn::computeCovar(), PLearn::GaussianDistribution::computeEigenDecomposition(), PLearn::RBMModule::computeEnergy(), PLearn::KLp0p1RBMModule::computeEnergy(), PLearn::RBMWoodsLayer::computeExpectations(), PLearn::RBMLateralBinomialLayer::computeExpectations(), PLearn::RBMModule::computeFreeEnergyOfHidden(), PLearn::KLp0p1RBMModule::computeFreeEnergyOfHidden(), PLearn::RBMModule::computeFreeEnergyOfVisible(), PLearn::KLp0p1RBMModule::computeFreeEnergyOfVisible(), PLearn::VMatKernel::computeGramMatrix(), PLearn::ThresholdedKernel::computeGramMatrix(), PLearn::RationalQuadraticARDKernel::computeGramMatrix(), PLearn::PLearnerDiagonalKernel::computeGramMatrix(), PLearn::Matern1ARDKernel::computeGramMatrix(), PLearn::Kernel::computeGramMatrix(), PLearn::IIDNoiseKernel::computeGramMatrix(), PLearn::Kernel::computeGramMatrixDerivative(), PLearn::IIDNoiseKernel::computeGramMatrixDerivative(), PLearn::RationalQuadraticARDKernel::computeGramMatrixDerivIspAlpha(), PLearn::SquaredExponentialARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::RationalQuadraticARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::NeuralNetworkARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::Matern1ARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::LinearARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::SquaredExponentialARDKernel::computeGramMatrixDerivIspSignalSigma(), PLearn::RationalQuadraticARDKernel::computeGramMatrixDerivIspSignalSigma(), PLearn::PLearnerDiagonalKernel::computeGramMatrixDerivIspSignalSigma(), PLearn::NeuralNetworkARDKernel::computeGramMatrixDerivIspSignalSigma(), PLearn::Matern1ARDKernel::computeGramMatrixDerivIspSignalSigma(), PLearn::LinearARDKernel::computeGramMatrixDerivIspSignalSigma(), PLearn::IIDNoiseKernel::computeGramMatrixDerivKronecker(), PLearn::MemoryCachedKernel::computeGramMatrixDerivNV(), PLearn::MemoryCachedKernel::computeGramMatrixNV(), PLearn::LayerCostModule::computeHisto(), PLearn::computeInputCovar(), PLearn::GaussMix::computeLogLikelihood(), PLearn::DeepNonLocalManifoldParzen::computeManifoldParzenParameters(), PLearn::computeMeanAndCovar(), PLearn::Kernel::computeNearestNeighbors(), PLearn::RBMModule::computeNegLogPVisibleGivenPHidden(), PLearn::ManifoldParzen2::computeOutput(), PLearn::ManifoldParzen::computeOutput(), PLearn::KNNClassifier::computeOutput(), PLearn::KernelProjection::computeOutput(), PLearn::BaggingLearner::computeOutput(), PLearn::ModuleLearner::computeOutputAndCosts(), PLearn::PLearner::computeOutputCovMat(), PLearn::GaussianProcessRegressor::computeOutputCovMat(), PLearn::StackedAutoassociatorsNet::computeOutputsAndCosts(), PLearn::PLearner::computeOutputsAndCosts(), PLearn::NatGradNNet::computeOutputsAndCosts(), PLearn::ModuleLearner::computeOutputsAndCosts(), PLearn::mNNet::computeOutputsAndCosts(), PLearn::DeepBeliefNet::computeOutputsAndCosts(), PLearn::RBMModule::computePartitionFunction(), PLearn::LayerCostModule::computePascalStatistics(), PLearn::RBMModule::computePositivePhaseHiddenActivations(), PLearn::KLp0p1RBMModule::computePositivePhaseHiddenActivations(), PLearn::GaussMix::computePosteriors(), PLearn::computePrincipalComponents(), PLearn::RBMSparse1DMatrixConnection::computeProducts(), PLearn::RBMMixedConnection::computeProducts(), PLearn::RBMMatrixTransposeConnection::computeProducts(), PLearn::RBMMatrixConnection::computeProducts(), PLearn::RBMDiagonalMatrixConnection::computeProducts(), PLearn::RBMConv2DConnection::computeProducts(), PLearn::computeRanks(), PLearn::LayerCostModule::computeSafeHisto(), PLearn::Kernel::computeSparseGramMatrix(), PLearn::LocalizedFeaturesLayerVariable::computeSubsets(), PLearn::UniformizeLearner::computeWeightedRankMap(), PLearn::ReconstructionWeightsKernel::computeWeights(), PLearn::SubsamplingDBN::contrastiveDivergenceStep(), PLearn::DeepBeliefNet::contrastiveDivergenceStep(), PLearn::ConvertFromPyObject< TMat< T > >::convert(), PLearn::ConvertFromPyObject< Mat >::convert(), PLearn::YMDDatedVMatrix::copyDatesOfRows(), PLearn::correlations(), PLearn::cross_valid(), PLearn::Variable::defineGradientLocation(), PLearn::Variable::defineValueLocation(), PLearn::det(), PLearn::DTWKernel::dtw(), PLearn::eigen_SymmMat(), PLearn::eigenSparseNonSymmMat(), PLearn::eigenSparseSymmMat(), PLearn::eigenVecOfSymmMat(), PLearn::MatlabInterface::eigs_r11(), PLearn::PCA::em_algo(), PLearn::PCA::em_orth_algo(), PLearn::ProductRandomVariable::EMTrainingInitialize(), PLearn::DenoisingRecurrentNet::encode_artificialData(), PLearn::DenoisingRecurrentNet::encode_onehot_diffNote_duration(), PLearn::DenoisingRecurrentNet::encode_onehot_note_octav_duration(), PLearn::DenoisingRecurrentNet::encode_onehot_timeframe(), PLearn::DenoisingRecurrentNet::encodeSequence(), PLearn::ConcatRowsVMatrix::ensureMappingsConsistency(), PLearn::GaussianProcessNLLVariable::fbpropFragments(), PLearn::LiftStatsCollector::finalize(), PLearn::BasisSelectionRegressor::findBestCandidateFunction(), PLearn::SubsamplingDBN::fineTuningStep(), PLearn::DeepNonLocalManifoldParzen::fineTuningStep(), PLearn::DeepBeliefNet::fineTuningStep(), PLearn::LiftStatsIterator::finish(), PLearn::VecStatsCollector::forget(), PLearn::ShuntingNNetLayerModule::forget(), PLearn::PseudolikelihoodRBM::forget(), PLearn::PLearnDiff::forget(), PLearn::ObservationWindow::forget(), PLearn::LLC::forget(), PLearn::LiftStatsCollector::forget(), PLearn::KMeansClustering::forget(), PLearn::KernelRidgeRegressor::forget(), PLearn::GradNNetLayerModule::forget(), PLearn::ExhaustiveNearestNeighbors::forget(), PLearn::Convolution2DModule::forget(), PLearn::BasisSelectionRegressor::forget(), PLearn::BackConvolution2DModule::forget(), PLearn::AutoLinearRegressor::forget(), PLearn::VBoundDBN2::fprop(), PLearn::TreeDBNModule::fprop(), PLearn::TanhModule::fprop(), PLearn::SquaredErrorCostModule::fprop(), PLearn::SplitModule::fprop(), PLearn::SoftSoftMaxVariable::fprop(), PLearn::SoftmaxNLLCostModule::fprop(), PLearn::SoftmaxModule::fprop(), PLearn::ShuntingNNetLayerModule::fprop(), PLearn::ScaleGradientModule::fprop(), PLearn::RBMWoodsLayer::fprop(), PLearn::RBMModule::fprop(), PLearn::RBMMixedLayer::fprop(), PLearn::RBMLayer::fprop(), PLearn::RBMLateralBinomialLayer::fprop(), PLearn::RBMConnection::fprop(), PLearn::RBMClassificationModule::fprop(), PLearn::RBMBinomialLayer::fprop(), PLearn::ProjectionErrorVariable::fprop(), PLearn::OnBagsModule::fprop(), PLearn::NllSemisphericalGaussianVariable::fprop(), PLearn::NllGeneralGaussianVariable::fprop(), PLearn::NLLCostModule::fprop(), PLearn::NetworkModule::fprop(), PLearn::MaxSubsampling2DModule::fprop(), PLearn::MatrixModule::fprop(), PLearn::LinearFilterModule::fprop(), PLearn::LinearCombinationModule::fprop(), PLearn::LayerCostModule::fprop(), PLearn::KLp0p1RBMModule::fprop(), PLearn::IdentityModule::fprop(), PLearn::GradNNetLayerModule::fprop(), PLearn::CrossEntropyCostModule::fprop(), PLearn::CostModule::fprop(), PLearn::Convolution2DModule::fprop(), PLearn::CombiningCostsModule::fprop(), PLearn::ClassErrorCostModule::fprop(), PLearn::BinarizeModule::fprop(), PLearn::ArgmaxModule::fprop(), PLearn::RBMMixedLayer::fpropNLL(), PLearn::SumOverBagsVariable::fpropOneBag(), PLearn::TreeDBNModule::full_fprop(), PLearn::GCV(), PLearn::generalizedEigenVecOfSymmMat(), PLearn::RBMDistribution::generate(), PLearn::DynamicallyLinkedRBMsModel::generate(), PLearn::DenoisingRecurrentNet::generate(), PLearn::DenoisingRecurrentNet::generateArtificial(), PLearn::RBMDistribution::generateN(), PLearn::GaussianContinuum::get_image_matrix(), PLearn::GaussianContinuum::get_nll(), PLearn::VecStatsCollector::getCovariance(), PLearn::VMatrix::getExamples(), PLearn::SplitModule::getPortSizes(), PLearn::OnlineLearningModule::getPortSizes(), PLearn::NullModule::getPortSizes(), PLearn::MatrixModule::getPortSizes(), PLearn::CostModule::getPortSizes(), PLearn::BinarizeModule::getPortSizes(), PLearn::DenoisingRecurrentNet::getSequence(), PLearn::SurfaceMesh::getVertexCoordsAndFeatures(), PLearn::OnlineGramNaturalGradientOptimizer::gramEigenNaturalGradient(), PLearn::SubsamplingDBN::greedyStep(), PLearn::DeepBeliefNet::greedyStep(), PLearn::LiftStatsIterator::init(), PLearn::NatGradEstimator::init(), PLearn::NeuralProbabilisticLanguageModel::initializeParams(), PLearn::FeatureSetSequentialCRF::initializeParams(), PLearn::FeatureSetNNet::initializeParams(), PLearn::TreeDBNModule::initSampling(), PLearn::TransformationLearner::initTransformsParameters(), PLearn::inverse(), PLearn::ICP::iterate(), PLearn::kernelPCAfromDotProducts(), PLearn::GaussMix::kmeans(), PLearn::NonLocalManifoldParzen::knn(), PLearn::GaussianContinuumDistribution::knn(), PLearn::GaussianContinuum::knn(), PLearn::lapackEIGEN(), PLearn::lapackGeneralizedEIGEN(), PLearn::lapackSVD(), PLearn::linearRegression(), PLearn::linearRegressionNoBias(), PLearn::RGBImageDB::load(), PLearn::loadAscii(), PLearn::loadAsciiSingleBinaryDescriptor(), PLearn::loadCorel(), PLearn::loadCorelDatamat(), PLearn::loadGnuplot(), PLearn::loadJPEGrgb(), PLearn::loadPMat(), PLearn::RBMDistribution::log_density(), PLearn::GaussianContinuumDistribution::log_density(), PLearn::TransformationLearner::mainLearnerBuild(), PLearn::matlabR11eigs(), PLearn::MemoryVMatrix::MemoryVMatrix(), PLearn::VecStatsCollector::merge(), PLearn::metricMultiDimensionalScaling(), PLearn::GaussMix::missingExpectation(), PLearn::Molecule::Molecule(), PLearn::multiply(), PLearn::OnlineLearningModule::namedBpropAccUpdate(), PLearn::SubsamplingDBN::onlineStep(), PLearn::StackedAutoassociatorsNet::onlineStep(), PLearn::DeepBeliefNet::onlineStep(), PLearn::BasisSelectionRegressor::thread_wawr::operator()(), PLearn::MaxSubsamplingTest::perform(), PLearn::PTester::perform1Split(), PLearn::ExhaustiveNearestNeighbors::preloadInputCache(), PLearn::Molecule::readMolecule(), PLearn::BasisSelectionRegressor::recomputeFeatures(), PLearn::ReconstructionWeightsKernel::reconstruct(), PLearn::DenoisingRecurrentNet::recurrentUpdate(), PLearn::regulargrid_x_y_rgbreal_to_bitmap(), PLearn::LinearRegressor::resetAccumulators(), PLearn::NnlmOutputLayer::resetAllClassVars(), PLearn::NnlmOutputLayer::resetParameters(), PLearn::UndirectedSoftmaxModule::resetWeights(), PLearn::NnlmWordRepresentationLayer::resetWeights(), PLearn::Variable::resize(), PLearn::DenoisingRecurrentNet::resize_lists(), PLearn::GaussMix::resizeDataBeforeTraining(), PLearn::GaussMix::resizeDataBeforeUsing(), PLearn::Variable::resizeDiagHessian(), PLearn::Variable::resizeRValue(), PLearn::BasisSelectionRegressor::retrainLearner(), PLearn::ridgeRegressionByGCV(), PLearn::ChemicalICP::run(), PLearn::scores_to_winners(), PLearn::RBMLayer::setBatchSize(), PLearn::ThresholdedKernel::setDataForKernelMatrix(), PLearn::MatrixModule::setGradientTo(), PLearn::GaussMix::setPredictor(), PLearn::GaussMix::setPredictorPredictedSizes_const(), PLearn::StackedFocusedAutoassociatorsNet::setTrainingSet(), PLearn::LocalGaussianClassifier::setTrainingSet(), PLearn::ExhaustiveNearestNeighbors::setTrainingSet(), PLearn::TransformationLearner::setTransformsParameters(), PLearn::StatsCollector::sort_values_by_magnitude(), PLearn::RegressionTreeRegisters::sortRows(), PLearn::SpearmanRankCorrelation(), PLearn::DenoisingRecurrentNet::splitRawMaskedSupervisedSequence(), PLearn::InferenceRBM::supCDStep(), PLearn::SVDVariable::SVDVariable(), PLearn::SymmMatNullSpaceByInversePowerIteration(), PLearn::PLearner::test(), PLearn::Learner::test(), PLearn::DynamicallyLinkedRBMsModel::test(), PLearn::DenoisingRecurrentNet::test(), PLearn::testSpearmanRankCorrelation(), PLearn::WPLS::train(), PLearn::VariableSelectionWithDirectedGradientDescent::train(), PLearn::SubsamplingDBN::train(), PLearn::StackedSVDNet::train(), PLearn::StackedAutoassociatorsNet::train(), PLearn::RankLearner::train(), PLearn::PseudolikelihoodRBM::train(), PLearn::PLS::train(), PLearn::NatGradSMPNNet::train(), PLearn::NatGradNNet::train(), PLearn::ModuleLearner::train(), PLearn::mNNet::train(), PLearn::ManifoldParzen::train(), PLearn::LinearRegressor::train(), PLearn::KMeansClustering::train(), PLearn::KernelRidgeRegressor::train(), PLearn::KernelProjection::train(), PLearn::IncrementalNNet::train(), PLearn::GaussMix::train(), PLearn::GaussianDistribution::train(), PLearn::GaussianContinuum::train(), PLearn::ExhaustiveNearestNeighbors::train(), PLearn::DynamicallyLinkedRBMsModel::train(), PLearn::DenoisingRecurrentNet::train(), PLearn::DeepNonLocalManifoldParzen::train(), PLearn::DeepBeliefNet::train(), PLearn::ConditionalGaussianDistribution::train(), PLearn::AutoLinearRegressor::train(), PLearn::ModuleLearner::trainingStep(), PLearn::DenoisingRecurrentNet::trainUnconditionalPredictor(), PLearn::transformPoints(), PLearn::TransformationLearner::treeDataSet(), PLearn::VecStatsCollector::update(), PLearn::LiftStatsIterator::update(), PLearn::RBMMatrixTransposeConnection::update(), PLearn::RBMMatrixConnection::update(), PLearn::RBMLLParameters::update(), PLearn::RBMConv2DLLParameters::update(), PLearn::RBMConv2DConnection::update(), PLearn::ObservationWindow::update(), PLearn::LiftStatsCollector::update(), PLearn::GaussianContinuumDistribution::update_reference_set_parameters(), PLearn::GaussianContinuum::update_reference_set_parameters(), PLearn::GaussMix::updateCholeskyFromPrevious(), PLearn::ToBagClassifier::updateCostAndBagOutput(), PLearn::RBMMatrixConnection::updateGibbs(), PLearn::GaussMix::updateInverseVarianceFromPrevious(), PLearn::DeepNonLocalManifoldParzen::updateManifoldParzenParameters(), PLearn::DiscriminativeDeepBeliefNet::updateNearestNeighbors(), PLearn::StackedFocusedAutoassociatorsNet::updateTrainSetRepresentations(), PLearn::PLearner::use(), PLearn::PLearner::useOnTrain(), PLearn::GaussianProcessRegressor::variance(), PLearn::NeuralProbabilisticLanguageModel::verify_gradient(), PLearn::FeatureSetSequentialCRF::verify_gradient(), PLearn::FeatureSetNNet::verify_gradient(), and PLearn::weightedRidgeRegressionByGCV().

    {
#ifdef BOUNDCHECK
        resizeBoundCheck(new_length, new_width);
#endif
        if (new_length==length_ && new_width==width_)
            return;
        else if(storage.isNull())
        {
            offset_ = 0;
            length_ = new_length;
            width_  = new_width;
            mod_    = new_width;
            long newsize=(long)length()*mod() + extra;
            storage = new Storage<T>(newsize);
        }
        else
        {
            int usage = storage->usage();
            if (usage > 1 && mod() != 0 && new_width > mod()-offset_%mod())
                resizeModError();
            else if (preserve_content && size() > 0)
                resizePreserve(new_length, new_width, extra);
            else {
                // 'new_size' takes into account the ABSOLUTELY REQUIRED size
                // to hold the elements of the matrix.  We only resize the
                // underlying storage when the latter is not big enough to hold
                // 'new_size'.  When resizing the storage, we include 'extra'
                // more elements to anticipate further resizes (e.g. coming
                // from appendRow()).  IMPORTANT NOTE: don't include those
                // 'extra' bytes in the computation of 'new_size', for
                // otherwise a matrix reallocation will occur EVERY TIME
                // appendRow is called, turning an amortized O(N) algorithm
                // into an O(N^2) one.
                long new_size = offset_+(long)new_length*MAX(mod(),new_width);
                if(new_size > storage->length())
                    storage->resize(new_size + extra);
                if(new_width > mod())
                    mod_ = new_width;
            }
            length_ = new_length;
            width_ = new_width;
        }
    }
template<class T >
void PLearn::TMat< T >::resizeBoundCheck ( int  new_length,
int  new_width 
) [inline, protected]

Perform bound-checking on resize.

Definition at line 446 of file TMat_impl.h.

References PLERROR.

{
    if(new_length<0 || new_width<0)
        PLERROR("IN TMat::resize(int new_length, int new_width)\nInvalid arguments (%d, %d)", new_length, new_width);
}
template<class T >
void PLearn::TMat< T >::resizeModError ( ) [protected]

Report PLERROR if we resize changing the mod with usage > 1.

Definition at line 453 of file TMat_impl.h.

References PLERROR.

{
    PLERROR("IN TMat::resize(int new_length, int new_width) - For safety "
            "reasons, increasing the width() beyond mod()-offset_ modulo "
            "mod() is not allowed when the storage is shared with others");
}
template<class T >
void PLearn::TMat< T >::resizePreserve ( int  new_length,
int  new_width,
int  extra = 0 
) [protected]

Utility function to resize a matrix while preserving contents.

Definition at line 386 of file TMat_impl.h.

References PLearn::max(), MAX, PLearn::min(), PLearn::sqrt(), PLearn::square(), PLearn::usage(), w, and x.

{
    int usage      = storage->usage();
    int new_size   = new_length*MAX(mod(),new_width);
    int new_offset = usage>1?offset_:0;
    if (new_size>storage->length() || new_width>mod())
    {
        int extracols=0, extrarows=0;
        if (extra>min(new_width,new_length))
        {
            // if width has increased, bet that it will increase again in the future,
            // similarly for length,  so allocate the extra as extra mod
            float l=float(length_), l1=float(new_length),
                w=float(width_),  w1=float(new_width),
                x=float(extra);
            // Solve the following equations to apportion the extra 
            // while keeping the same percentage increase in width and length:
            //   Solve[{x+w1*l1==w2*l2,(w2/w1 - 1)/(l2/l1 - 1) == (w1/w - 1)/(l1/l - 1)},{w2,l2}]
            // This is a quadratic system which has two solutions: {w2a,l2a} and {w2b,l2b}:
            float w2a = 
                w1*(-1 - l1/(l - l1) + w1/w + (l1*w1)/(l*w - l1*w) + 
                    (2*l*(-w + w1)*x)/
                    (2*l*l1*w*w1 - l1*l1*w*w1 - l*l1*w1*w1 + 
                     sqrt(square(l1*l1*w*w1 - l*l1*w1*w1) + 
                          4*l*(l - l1)*l1*w*(w - w1)*w1*(l1*w1 + x))));
            float l2a = -(-l1*l1*w*w1 + l*l1*w1*w1 + 
                          sqrt(square(l1*l1*w*w1 - l*l1*w1*w1) + 
                               4*l*(l - l1)*l1*w*(w - w1)*w1*(l1*w1 + x)))/(2*l*(w - w1)*w1);
            float w2b =w1*(-1 - l1/(l - l1) + w1/w + (l1*w1)/(l*w - l1*w) - 
                           (2*l*(-w + w1)*x)/
                           (-2*l*l1*w*w1 + l1*l1*w*w1 + l*l1*w1*w1 + 
                            sqrt(square(l1*l1*w*w1 - l*l1*w1*w1) + 
                                 4*l*(l - l1)*l1*w*(w - w1)*w1*(l1*w1 + x))));
            float l2b = (l1*l1*w*w1 - l*l1*w1*w1 + 
                         sqrt(square(l1*l1*w*w1 - l*l1*w1*w1) + 
                              4*l*(l - l1)*l1*w*(w - w1)*w1*(l1*w1 + x)))/(2*l*(w - w1)*w1);

            // pick one that is feasible and maximizes the mod
            if (w2b>w2a && w2b>w1 && l2b>l1) {
                extracols=int(ceil(w2b-w1));
                extrarows=int(ceil(l2b-l1));
            }
            else if (w2a>w1 && l2a>l1) {
                extrarows=int(ceil(l2a-l1));
                extracols=int(ceil(w2a-w1));
            }
            else { // no valid solution to the system of equation, use a heuristic
                extracols = max(0,int(ceil(sqrt(real(extra))/new_length)));
                extrarows = max(0,int((extra+l1*w1)/(w1+extracols) - l1));
            }

        }
        storage->resizeMat(new_length,new_width,extrarows,extracols,
                           new_offset,mod_,length_,width_,offset_);
        mod_ = new_width + extracols;
    }
    offset_ = new_offset;
}

Here is the call graph for this function:

template<class T>
TMat<T> PLearn::TMat< T >::row ( int  row) const [inline]

Returns a Mat that is a row of the matrix.

Definition at line 590 of file TMat_decl.h.

Referenced by PLearn::VarRowsVariable::bprop(), PLearn::KNNVMatrix::build_(), PLearn::OnBagsModule::fprop(), and PLearn::GaussianProcessNLLVariable::fprop().

    { return subMatRows(row, 1); }

Here is the caller graph for this function:

template<class T>
T* PLearn::TMat< T >::rowdata ( int  i) const [inline]
template<class T>
rowelements_iterator PLearn::TMat< T >::rowelements_begin ( int  rownum) const [inline]

returns an iterator over the elements of a particular row

Definition at line 147 of file TMat_decl.h.

    {
#ifdef BOUNDCHECK
        if(rownum<0 || rownum>=length())
            PLERROR("OUT OF RANGE rownum in rowelements_begin");
#endif
        return data()+rownum*mod();
    }
template<class T>
rowelements_iterator PLearn::TMat< T >::rowelements_end ( int  rownum) const [inline]

IMPORTANT WARNING: use this only to check reaching the end with an iterator obtained through rowelements_begin USING THE *SAME* rownum.

Definition at line 158 of file TMat_decl.h.

    { return data()+rownum*mod()+width(); }
template<class T>
template<class I >
TMat<T> PLearn::TMat< T >::rows ( const TVec< I > &  rows) const [inline]

selectRows(*this,rows,result) i.e.

return the matrix with specified rows (indices)

Definition at line 614 of file TMat_decl.h.

    {
        TMat<T> result(rows.length(),width());
        selectRows(*this,rows,result);
        return result;
    }
template<class T >
TMatRowsAsArraysIterator< T > PLearn::TMat< T >::rows_as_arrays_begin ( )

Return an iterator over all rows of the matrix.

No const version for now

Definition at line 503 of file TMat_impl.h.

Referenced by PLearn::sortRows().

                                                          {
    return TMatRowsAsArraysIterator<T>(data(), width_, mod_);
}

Here is the caller graph for this function:

template<class T >
TMatRowsAsArraysIterator< T > PLearn::TMat< T >::rows_as_arrays_end ( )

Definition at line 508 of file TMat_impl.h.

Referenced by PLearn::sortRows().

                                                        {
    return TMatRowsAsArraysIterator<T>(data()+length_*mod_, width_, mod_);
}

Here is the caller graph for this function:

template<class T >
TMatRowsIterator< T > PLearn::TMat< T >::rows_begin ( )

Return an iterator over all rows of the matrix.

No const version for now

Definition at line 492 of file TMat_impl.h.

                                        {
    return TMatRowsIterator<T>(data(), width_, mod_);
}
template<class T >
TMatRowsIterator< T > PLearn::TMat< T >::rows_end ( )

Definition at line 497 of file TMat_impl.h.

                                      {
    return TMatRowsIterator<T>(data()+length_*mod_, width_, mod_);
}
template<class T>
void PLearn::TMat< T >::setMod ( int  new_mod) [inline]

Set a new value for 'mod'.

The content of the matrix will be destroyed (i.e. moved around). In addition, if the new mod is strictly less than the width, the width will be set to this new mod (in order to ensure it remains a valid Mat).

Definition at line 271 of file TMat_decl.h.

Referenced by PLearn::GaussMix::addToCovariance(), PLearn::GaussMix::computeLogLikelihood(), PLearn::Variable::defineGradientLocation(), PLearn::Variable::defineValueLocation(), PLearn::Variable::resize(), PLearn::GaussMix::setPredictor(), and PLearn::GaussMix::updateInverseVarianceFromPrevious().

    {
        if (new_mod == mod())
            // Nothing to do (the new mod is equal to the old one).
            return;
        if (storage.isNull()) {
            mod_ = new_mod;
            return;
        }
        if (storage->usage() > 1)
            PLERROR("In setMod - You cannot change the 'mod' of a matrix "
                    "whose storage is shared");
        if (new_mod > mod()) {
            // The mod is increased: we may need a larger storage. To this
            // extent, the matrix is first resized to a width equal to the new
            // mod, to ensure the storage is large enough for the new mod.
            int width_backup = width();
            resize(length(), new_mod);
            PLASSERT( mod() == new_mod );
            width_ = width_backup;
        } else {
            // Note that since new_mod < curent mod, then the storage is
            // necessarily already large enough and does not need resizing.
            mod_ = new_mod;
            if (new_mod < width()) {
                // We cannot just change the mod, because in order to be a
                // valid Mat, we must have mod >= width. Thus we also change
                // the width to match the new mod.
                width_ = new_mod;
            }
        }
    }

Here is the caller graph for this function:

template<class T>
int PLearn::TMat< T >::size ( ) const [inline]

Definition at line 251 of file TMat_decl.h.

Referenced by PLearn::UndirectedSoftmaxModule::build_(), PLearn::NnlmWordRepresentationLayer::build_(), PLearn::NnlmOutputLayer::build_(), PLearn::LinearInductiveTransferClassifier::build_(), PLearn::Convolution2DModule::build_kernels(), PLearn::BackConvolution2DModule::build_kernels(), PLearn::SquaredExponentialARDKernel::computeGramMatrix(), PLearn::RationalQuadraticARDKernel::computeGramMatrix(), PLearn::PLearnerDiagonalKernel::computeGramMatrix(), PLearn::NeuralNetworkARDKernel::computeGramMatrix(), PLearn::Matern1ARDKernel::computeGramMatrix(), PLearn::LinearARDKernel::computeGramMatrix(), PLearn::KroneckerBaseKernel::computeGramMatrix(), PLearn::MemoryCachedKernel::dataCached(), PLearn::dot(), PLearn::fill_one_hot(), PLearn::ExhaustiveNearestNeighbors::findNearestNeighbors(), PLearn::UndirectedSoftmaxModule::forget(), PLearn::PseudolikelihoodRBM::forget(), PLearn::GradNNetLayerModule::forget(), PLearn::RBMModule::fprop(), PLearn::DenoisingRecurrentNet::inject_zero_forcing_noise(), PLearn::logadd(), PLearn::RBMQLParameters::makeParametersPointHere(), PLearn::RBMMatrixConnection::makeParametersPointHere(), PLearn::RBMLQParameters::makeParametersPointHere(), PLearn::RBMLLParameters::makeParametersPointHere(), PLearn::RBMGenericParameters::makeParametersPointHere(), PLearn::RBMConv2DLLParameters::makeParametersPointHere(), PLearn::RBMConv2DConnection::makeParametersPointHere(), PLearn::max(), PLearn::maxabs(), PLearn::min(), PLearn::minabs(), PLearn::multinomial_sample(), PLearn::RBMSparse1DMatrixConnection::nParameters(), PLearn::RBMQLParameters::nParameters(), PLearn::RBMMatrixTransposeConnection::nParameters(), PLearn::RBMMatrixConnection::nParameters(), PLearn::RBMLQParameters::nParameters(), PLearn::RBMLLParameters::nParameters(), PLearn::RBMGenericParameters::nParameters(), PLearn::RBMConv2DLLParameters::nParameters(), PLearn::RBMConv2DConnection::nParameters(), performLP(), PLearn::SurfaceMesh::readVRMLCoordinate3_(), PLearn::sizeInBytes(), PLearn::squareElements(), PLearn::sumabs(), PLearn::sumsquare(), and PLearn::HistogramDistribution::variance().

    { return length_*width_; }
template<class T>
TMat<T> PLearn::TMat< T >::subMat ( int  rowstart,
int  colstart,
int  newlength,
int  newwidth 
) const [inline]

Returns a sub-matrix that is a rectangular portion of this matrix.

Definition at line 633 of file TMat_decl.h.

Referenced by PLearn::VecStatsCollector::append(), PLearn::MaxSubsampling2DModule::bpropAccUpdate(), PLearn::RBMLocalMultinomialLayer::bpropUpdate(), PLearn::StackedSVDNet::build_(), PLearn::ProjectionErrorVariable::build_(), PLearn::chol_rotapp_tr_opt(), PLearn::choleskyRemoveDimension(), PLearn::DeepReconstructorNet::computeAndSaveLayerActivationStats(), PLearn::DeepBeliefNet::computeClassifAndFinalCostsFromOutputs(), PLearn::RBMLocalMultinomialLayer::computeExpectation(), PLearn::RBMLocalMultinomialLayer::computeExpectations(), PLearn::GaussMix::computeLogLikelihood(), PLearn::ReconstructionWeightsKernel::evaluate_x_i_again(), PLearn::StackedSVDNet::fineTuningStep(), PLearn::LiftStatsIterator::finish(), PLearn::RunICPVariable::fprop(), PLearn::RBMLocalMultinomialLayer::fprop(), PLearn::MaxSubsampling2DModule::fprop(), PLearn::RBMLocalMultinomialLayer::fpropNLL(), PLearn::RBMLocalMultinomialLayer::freeEnergyContribution(), PLearn::RBMLocalMultinomialLayer::generateSample(), PLearn::RBMLocalMultinomialLayer::generateSamples(), PLearn::RBMLocalMultinomialLayer::getConfiguration(), PLearn::VMatrixFromDistribution::getMat(), PLearn::MemoryVMatrix::getMat(), PLearn::SplitModule::getPortSizes(), PLearn::OnlineGramNaturalGradientOptimizer::gramEigenNaturalGradient(), PLearn::NatGradEstimator::init(), PLearn::StatsCollector::lift(), PLearn::loadCorel(), PLearn::StatsCollector::prbp(), PLearn::MemoryVMatrix::putMat(), PLearn::rotationFromWeightedMatchedPoints(), PLearn::GaussMix::setPredictor(), PLearn::GaussMix::setPredictorPredictedSizes_const(), PLearn::Variable::setValueSubMat(), PLearn::MemoryVMatrix::subMat(), PLearn::testCholeskyRoutines(), PLearn::WPLS::train(), PLearn::StackedSVDNet::train(), PLearn::SequentialModelSelector::train(), PLearn::GaussMix::updateInverseVarianceFromPrevious(), and PLearn::weightedRotationFromMatchedPoints().

    {
#ifdef BOUNDCHECK
        if(rowstart<0 || newlength<0 || rowstart+newlength>length()
           || colstart<0 || newwidth<0 || colstart+newwidth>width())
            PLERROR("Mat::subMat(int rowstart, int colstart, int newlength, int newwidth) OUT OF BOUNDS"
                    "  rowstart=%d colstart=%d newlength=%d newwidth=%d length()=%d width()=%d",
                    rowstart, colstart, newlength, newwidth, length(), width());
#endif
        TMat<T> subm = *this;
        subm.length_ = newlength;
        subm.width_ = newwidth;
        subm.offset_ += rowstart*mod() + colstart;
        return subm;
    }

Here is the caller graph for this function:

template<class T>
TMat<T> PLearn::TMat< T >::subMatColumns ( int  colstart,
int  newwidth 
) const [inline]

Returns a sub-matrix that is a range of columns of this matrix.

Definition at line 664 of file TMat_decl.h.

Referenced by PLearn::RBMSparse1DMatrixConnection::accumulateNegStats(), PLearn::RBMSparse1DMatrixConnection::accumulatePosStats(), PLearn::StringTable::appendRow(), PLearn::SplitModule::bpropAccUpdate(), PLearn::RBMSparse1DMatrixConnection::bpropAccUpdate(), PLearn::RBMMixedConnection::bpropAccUpdate(), PLearn::RBMMixedLayer::bpropNLL(), PLearn::RBMSparse1DMatrixConnection::bpropUpdate(), PLearn::RBMMixedLayer::bpropUpdate(), PLearn::RBMMixedConnection::bpropUpdate(), PLearn::RBMJointLLParameters::build_(), PLearn::RBMJointGenericParameters::build_(), PLearn::NatGradSMPNNet::build_(), PLearn::NatGradNNet::build_(), PLearn::mNNet::build_(), PLearn::chol_dxch_tr(), PLearn::RGBImageDB::computeHistogramRepresentation(), PLearn::RBMJointGenericParameters::computeLinearUnitActivations(), PLearn::ModuleLearner::computeOutputsAndCosts(), PLearn::AddCostToLearner::computeOutputsAndCosts(), PLearn::RBMMatrixTransposeConnection::computeProduct(), PLearn::RBMMatrixConnection::computeProduct(), PLearn::RBMSparse1DMatrixConnection::computeProducts(), PLearn::RBMMixedConnection::computeProducts(), PLearn::RBMMatrixTransposeConnection::computeProducts(), PLearn::RBMMatrixConnection::computeProducts(), PLearn::RBMJointLLParameters::computeUnitActivations(), PLearn::PCA::em_orth_algo(), PLearn::DenoisingRecurrentNet::encode_artificialData(), PLearn::Kernel::estimateHistograms(), PLearn::GaussianProcessNLLVariable::fbpropFragments(), PLearn::Gnuplot::featureplot(), PLearn::SubsamplingDBN::fineTuningStep(), PLearn::DeepBeliefNet::fineTuningStep(), PLearn::SplitModule::fprop(), PLearn::RBMMixedLayer::fprop(), PLearn::CombiningCostsModule::fprop(), PLearn::RBMMixedLayer::fpropNLL(), PLearn::OnlineGramNaturalGradientOptimizer::gramEigenNaturalGradient(), PLearn::hconcat(), PLearn::InferenceRBM::hiddenExpGivenInput(), PLearn::ICP::iterate(), PLearn::loadATT800(), PLearn::loadBreastCancerWisconsin(), PLearn::loadDiabetes(), PLearn::loadHousing(), PLearn::loadLetters(), PLearn::loadPimaIndians(), PLearn::loadUCI(), PLearn::newIndexedMatArray(), PLearn::normalizeDataSets(), PLearn::SubsamplingDBN::onlineStep(), PLearn::StackedAutoassociatorsNet::onlineStep(), PLearn::DeepBeliefNet::onlineStep(), PLearn::operator^(), PLearn::MoleculeTemplate::readFromAMATFile(), PLearn::regulargrid_x_y_outputs_to_bitmap(), PLearn::removeColumn(), PLearn::RBMMixedConnection::setAsDownInputs(), PLearn::RBMMixedConnection::setAsUpInputs(), PLearn::RBMMixedLayer::setExpectationsByRef(), PLearn::DenoisingRecurrentNet::splitRawMaskedSupervisedSequence(), PLearn::testCholeskyRoutines(), PLearn::AutoLinearRegressor::train(), PLearn::ModuleLearner::trainingStep(), PLearn::BasicIdentityCallsTest::unary(), PLearn::RBMSparse1DMatrixConnection::update(), PLearn::RBMMixedLayer::update(), and PLearn::RBMMixedConnection::update().

    {
#ifdef BOUNDCHECK
        if(colstart<0 || newwidth<0 || colstart+newwidth>width())
            PLERROR("Mat::subMatColumns(int colstart, int newwidth) OUT OF BOUNDS");
#endif
        TMat<T> subm = *this;
        subm.width_ = newwidth;
        subm.offset_ += colstart;
        return subm;
    }
template<class T>
TMat<T> PLearn::TMat< T >::subMatRows ( int  rowstart,
int  newlength 
) const [inline]

Returns a sub-matrix that is a range of rows of this matrix.

Definition at line 650 of file TMat_decl.h.

Referenced by PLearn::SparseIncrementalAffineTransformVariable::bprop(), PLearn::MatrixAffineTransformVariable::bprop(), PLearn::AffineTransformVariable::bprop(), PLearn::OnBagsModule::bpropUpdate(), PLearn::TransformationLearner::buildLearnedParameters(), PLearn::chol_dxch(), PLearn::Kernel::computeKNNeighbourMatrixFromDistanceMatrix(), PLearn::KernelProjection::computeOutput(), PLearn::IsomapTangentLearner::computeOutput(), PLearn::BaggingLearner::computeOutput(), PLearn::RBMMatrixTransposeConnection::computeProduct(), PLearn::RBMMatrixConnection::computeProduct(), PLearn::RBMMatrixTransposeConnection::computeProducts(), PLearn::RBMMatrixConnection::computeProducts(), PLearn::diagonalizeSubspace(), PLearn::PCA::em_orth_algo(), PLearn::DenoisingRecurrentNet::encodeAndCreateSupervisedSequence(), PLearn::DenoisingRecurrentNet::encodeAndCreateSupervisedSequence2(), PLearn::NatGradSMPNNet::fbpropLoss(), PLearn::NatGradNNet::fbpropLoss(), PLearn::mNNet::fbpropLoss(), PLearn::MatrixAffineTransformVariable::fprop(), PLearn::MatrixAffineTransformFeedbackVariable::fprop(), PLearn::NatGradSMPNNet::fpropNet(), PLearn::NatGradNNet::fpropNet(), PLearn::mNNet::fpropNet(), PLearn::RBMDistribution::generateN(), PLearn::TransformationLearner::generatorBuild(), PLearn::DenoisingRecurrentNet::getInputWindow(), PLearn::GramSchmidtOrthogonalization(), PLearn::TransformationLearner::initTransformsParameters(), PLearn::loadATT800(), PLearn::loadBreastCancer(), PLearn::loadDiabetes(), PLearn::loadLetters(), PLearn::loadUCI(), PLearn::TransformationLearner::mainLearnerBuild(), PLearn::newIndexedMatArray(), PLearn::SourceVariable::randomInitialize(), PLearn::removeRow(), PLearn::TestDependenciesCommand::run(), PLearn::LocalGaussianClassifier::setTrainingSet(), PLearn::TransformationLearner::setTransformsParameters(), PLearn::SymmMatNullSpaceByInversePowerIteration(), PLearn::LLC::train(), PLearn::LinearRegressor::train(), PLearn::KernelProjection::train(), PLearn::GaussianProcessRegressor::trainProjectedProcess(), PLearn::TransformationLearner::treeDataSet(), and PLearn::vconcat().

    {
#ifdef BOUNDCHECK
        if(rowstart<0 || newlength<0 || rowstart+newlength>length())
            PLERROR("TMat::subMatRows(int rowstart, int newlength) OUT OF BOUNDS"
                    "length=%d, rowstart=%d, newlength=%d", length(), rowstart, newlength);
#endif
        TMat<T> subm = *this;
        subm.length_ = newlength;
        subm.offset_ += rowstart*mod();
        return subm;
    }
template<class T>
void PLearn::TMat< T >::swapColumns ( int  i,
int  j 
) const [inline]

Swap the content of column i and column j.

Definition at line 811 of file TMat_decl.h.

Referenced by PLearn::chol_dxch().

    {
        if (i != j)
        {
            T* Mi = data() + i;
            T* Mj = data() + j;
            int n = length();
            for (int k = 0; k < n; k++) {
                T tmp = *Mi;
                *Mi = *Mj;
                *Mj = tmp;
                Mi += mod();
                Mj += mod();
            }
        }
    }

Here is the caller graph for this function:

template<class T>
void PLearn::TMat< T >::swapRows ( int  i,
int  j 
) const [inline]

Swap the content of row i and row j.

Note: a potentially more efficient version can be found in TMat_maths_impl.h.

Definition at line 793 of file TMat_decl.h.

Referenced by PLearn::LU_decomposition(), PLearn::partialSortRows(), PLearn::selectAndOrder(), PLearn::shuffleRows(), PLearn::PRandom::shuffleRows(), PLearn::sortRows(), and PLearn::swapRows().

    {
        if(i!=j)
        {
            //T* Mi = rowdata(i);
            //T* Mj = rowdata(j);
            T* Mi = (*this)[i];
            T* Mj = (*this)[j];
            for (int k=0;k<width();k++)
            {
                T tmp = Mi[k];
                Mi[k] = Mj[k];
                Mj[k] = tmp;
            }
        }
    }

Here is the caller graph for this function:

template<class T>
void PLearn::TMat< T >::swapUpsideDown ( ) const [inline]

Definition at line 933 of file TMat_decl.h.

Referenced by PLearn::eigen_SymmMat_decreasing(), PLearn::eigenVecOfSymmMat(), PLearn::generalizedEigenVecOfSymmMat(), and PLearn::kernelPCAfromDotProducts().

    {
        int half = length()/2;
        for(int i=0; i<half; i++)
            swapRows(i, length()-i-1);
    }

Here is the caller graph for this function:

template<class T >
TVec< T > PLearn::TMat< T >::toVec ( ) const

Views same data (not always possible)

Views same data (not always possible) Actually it's the matrix view rows by rows.

Definition at line 302 of file TMat_impl.h.

References PLearn::TVec< T >::length_, PLearn::TVec< T >::offset_, PLERROR, and PLearn::TVec< T >::storage.

Referenced by PLearn::VBoundDBN2::bpropAccUpdate(), PLearn::RBMSparse1DMatrixConnection::bpropAccUpdate(), PLearn::CostModule::bpropAccUpdate(), PLearn::RBMSparse1DMatrixConnection::bpropUpdate(), PLearn::SumOverBagsVariable::build_(), PLearn::RBMModule::computeLogLikelihoodOfVisible(), PLearn::LinearRegressor::computeOutput(), PLearn::KPCATangentLearner::computeOutput(), PLearn::IsomapTangentLearner::computeOutput(), PLearn::BaggingLearner::computeOutput(), PLearn::NNet::computeOutputAndCosts(), PLearn::RBMSparse1DMatrixConnection::computeProducts(), PLearn::ConvertFromPyObject< TMat< T > >::convert(), PLearn::ConvertFromPyObject< Mat >::convert(), PLearn::Variable::defineGradientLocation(), PLearn::Variable::defineValueLocation(), PLearn::ProductRandomVariable::EMTrainingInitialize(), PLearn::DenoisingRecurrentNet::encodeAndCreateSupervisedSequence(), PLearn::DenoisingRecurrentNet::encodeAndCreateSupervisedSequence2(), PLearn::GCV(), PLearn::RBMDistribution::generate(), PLearn::RBMDistribution::generateN(), PLearn::DenoisingRecurrentNet::getInputWindow(), PLearn::NeuralProbabilisticLanguageModel::initializeParams(), PLearn::FeatureSetSequentialCRF::initializeParams(), PLearn::FeatureSetNNet::initializeParams(), PLearn::NonLocalManifoldParzen::log_density(), PLearn::multivariate_normal(), PLearn::Variable::resize(), PLearn::Variable::resizeDiagHessian(), PLearn::Variable::resizeRValue(), PLearn::BinSplitter::setDataSet(), PLearn::NonLocalManifoldParzen::train(), PLearn::GaussianProcessRegressor::train(), PLearn::MemoryStressTest::unary(), PLearn::BasicIdentityCallsTest::unary(), PLearn::RBMSparse1DMatrixConnection::update(), PLearn::ConditionalDistribution::use(), PLearn::Variable::Variable(), PLearn::NeuralProbabilisticLanguageModel::verify_gradient(), PLearn::FeatureSetSequentialCRF::verify_gradient(), and PLearn::FeatureSetNNet::verify_gradient().

{
    if(length()>1 && width()<mod())
        PLERROR("In Mat::toVec internal structure of this Mat makes it impossible to build a Vec that would view exactly the same data. Consider using toVecCopy() instead!");
 
    TVec<T> v;
    v.offset_ = offset_;
    v.length_ = length()*width();
    v.storage = storage;
    return v;
}

Here is the caller graph for this function:

template<class T >
TVec< T > PLearn::TMat< T >::toVecCopy ( ) const
template<class T>
void PLearn::TMat< T >::transpose ( ) [inline]

Swap element (i,j) with element (j,i).

Currently only implemented for square matrices.

Definition at line 919 of file TMat_decl.h.

    {
        if (length() != width())
            PLERROR("In TMat<T>::tranpose() - Only implemented for square "
                    "matrices");
        for (int i = 0; i < length(); i++)
        {
            T* rowi = (*this)[i] + i + 1;
            T* colielem = rowi - 1 + mod();
            for(int j = i + 1; j < width(); j++, colielem += mod(), rowi++)
                pl_swap(*rowi, *colielem);
        }
    }
template<class T>
int PLearn::TMat< T >::width ( ) const [inline]

Definition at line 248 of file TMat_decl.h.

Referenced by PLearn::absargmax(), PLearn::add(), PLearn::RBMLayer::addBiasDecay(), PLearn::addToColumns(), PLearn::GaussMix::addToCovariance(), PLearn::RBMMatrixConnection::addWeightPenalty(), PLearn::affineMatrixInitialize(), PLearn::affineNormalization(), PLearn::GaussianProcessNLLVariable::alpha(), PLearn::VecStatsCollector::append(), PLearn::apply(), PLearn::argmax(), PLearn::argmin(), PLearn::averageAcrossRowsAndColumns(), PLearn::backConvolve2D(), PLearn::backConvolve2Dbackprop(), PLearn::TransformationLearner::batchGeneratePredictedFrom(), PLearn::bootstrap_rows(), PLearn::TransposedDoubleProductVariable::bprop(), PLearn::TraceVariable::bprop(), PLearn::SumVarianceOfLinearTransformedBernoullis::bprop(), PLearn::SumEntropyOfCategoricals::bprop(), PLearn::SumEntropyOfBernoullis::bprop(), PLearn::SparseIncrementalAffineTransformVariable::bprop(), PLearn::NonDiagVariable::bprop(), PLearn::LogSoftSoftMaxVariable::bprop(), PLearn::InsertZerosVariable::bprop(), PLearn::GaussianProcessNLLVariable::bprop(), PLearn::DoubleProductVariable::bprop(), PLearn::Cov2CorrVariable::bprop(), PLearn::ConstrainVariable::bprop(), PLearn::NllGeneralGaussianVariable::bprop_to_bases(), PLearn::DeepNonLocalManifoldParzen::bprop_to_bases(), PLearn::TreeDBNModule::bpropAccUpdate(), PLearn::SplitModule::bpropAccUpdate(), PLearn::SoftmaxNLLCostModule::bpropAccUpdate(), PLearn::RBMSparse1DMatrixConnection::bpropAccUpdate(), PLearn::RBMModule::bpropAccUpdate(), PLearn::RBMMixedConnection::bpropAccUpdate(), PLearn::RBMMatrixConnection::bpropAccUpdate(), PLearn::RBMConv2DConnection::bpropAccUpdate(), PLearn::OnlineLearningModule::bpropAccUpdate(), PLearn::OnBagsModule::bpropAccUpdate(), PLearn::NLLCostModule::bpropAccUpdate(), PLearn::NetworkModule::bpropAccUpdate(), PLearn::MaxSubsampling2DModule::bpropAccUpdate(), PLearn::MatrixModule::bpropAccUpdate(), PLearn::LinearCombinationModule::bpropAccUpdate(), PLearn::LayerCostModule::bpropAccUpdate(), PLearn::KLp0p1RBMModule::bpropAccUpdate(), PLearn::IdentityModule::bpropAccUpdate(), PLearn::CrossEntropyCostModule::bpropAccUpdate(), PLearn::CostModule::bpropAccUpdate(), PLearn::Convolution2DModule::bpropAccUpdate(), PLearn::CombiningCostsModule::bpropAccUpdate(), PLearn::BinarizeModule::bpropAccUpdate(), PLearn::mNNet::bpropNet(), PLearn::RBMWoodsLayer::bpropNLL(), PLearn::RBMMultinomialLayer::bpropNLL(), PLearn::RBMMixedLayer::bpropNLL(), PLearn::RBMLocalMultinomialLayer::bpropNLL(), PLearn::RBMLateralBinomialLayer::bpropNLL(), PLearn::RBMGaussianLayer::bpropNLL(), PLearn::RBMBinomialLayer::bpropNLL(), PLearn::SquaredErrorCostModule::bpropUpdate(), PLearn::SoftmaxNLLCostModule::bpropUpdate(), PLearn::SoftmaxModule::bpropUpdate(), PLearn::ShuntingNNetLayerModule::bpropUpdate(), PLearn::ScaleGradientModule::bpropUpdate(), PLearn::RBMWoodsLayer::bpropUpdate(), PLearn::RBMSparse1DMatrixConnection::bpropUpdate(), PLearn::RBMMultinomialLayer::bpropUpdate(), PLearn::RBMMixedLayer::bpropUpdate(), PLearn::RBMMixedConnection::bpropUpdate(), PLearn::RBMMatrixTransposeConnection::bpropUpdate(), PLearn::RBMMatrixConnectionNatGrad::bpropUpdate(), PLearn::RBMMatrixConnection::bpropUpdate(), PLearn::RBMLocalMultinomialLayer::bpropUpdate(), PLearn::RBMLateralBinomialLayer::bpropUpdate(), PLearn::RBMDiagonalMatrixConnection::bpropUpdate(), PLearn::RBMConv2DConnection::bpropUpdate(), PLearn::RBMBinomialLayer::bpropUpdate(), PLearn::ProcessInputCostModule::bpropUpdate(), PLearn::OnlineLearningModule::bpropUpdate(), PLearn::OnBagsModule::bpropUpdate(), PLearn::NLLCostModule::bpropUpdate(), PLearn::ModuleStackModule::bpropUpdate(), PLearn::LinearFilterModule::bpropUpdate(), PLearn::LayerCostModule::bpropUpdate(), PLearn::GradNNetLayerModule::bpropUpdate(), PLearn::CrossEntropyCostModule::bpropUpdate(), PLearn::CombiningCostsModule::bpropUpdate(), PLearn::mNNet::bpropUpdateNet(), PLearn::VMatrixFromDistribution::build_(), PLearn::SplitWiseValidationVMatrix::build_(), PLearn::ShuntingNNetLayerModule::build_(), PLearn::RemapLastColumnVMatrix::build_(), PLearn::RBMSparse1DMatrixConnection::build_(), PLearn::RBMQLParameters::build_(), PLearn::RBMMatrixConnection::build_(), PLearn::RBMLQParameters::build_(), PLearn::RBMLLParameters::build_(), PLearn::RBMGenericParameters::build_(), PLearn::RBMConv2DLLParameters::build_(), PLearn::RBMConv2DConnection::build_(), PLearn::ObjectOptionVariable::build_(), PLearn::MoleculeTemplate::build_(), PLearn::ModuleTester::build_(), PLearn::MemoryVMatrix::build_(), PLearn::LinearInductiveTransferClassifier::build_(), PLearn::LemmatizeVMatrix::build_(), PLearn::GradNNetLayerModule::build_(), PLearn::GaussMix::build_(), PLearn::FNetLayerVariable::build_(), PLearn::ExtractNNetParamsVMatrix::build_(), PLearn::Convolution2DModule::build_(), PLearn::ClassErrorCostModule::build_(), PLearn::BackConvolution2DModule::build_(), PLearn::AppendNeighborsVMatrix::build_(), PLearn::Convolution2DModule::build_kernels(), PLearn::BackConvolution2DModule::build_kernels(), PLearn::ChemicalICP::cacheFeatureDistances(), PLearn::chol_dxch_tr(), PLearn::chol_rotapp_tr(), PLearn::choleskyAppendDimension(), PLearn::choleskyDecomposition(), PLearn::choleskyLeftSolve(), PLearn::choleskyRightSolve(), PLearn::choleskySolve(), PLearn::classification_confusion_matrix(), PLearn::columnArgmax(), PLearn::columnArgmin(), PLearn::columnMax(), PLearn::columnMean(), PLearn::columnMin(), PLearn::columnSum(), PLearn::columnSumOfSquares(), PLearn::columnVariance(), PLearn::columnWeightedMean(), PLearn::columnWeightedVariance(), PLearn::GaussianContinuum::compute_train_and_validation_costs(), PLearn::DeepBeliefNet::computeClassifAndFinalCostsFromOutputs(), PLearn::AddCostToLearner::computeCostsFromOutputs(), PLearn::computeCovar(), PLearn::RBMWoodsLayer::computeExpectations(), PLearn::RBMTruncExpLayer::computeExpectations(), PLearn::RBMRateLayer::computeExpectations(), PLearn::RBMMultinomialLayer::computeExpectations(), PLearn::RBMLocalMultinomialLayer::computeExpectations(), PLearn::RBMLateralBinomialLayer::computeExpectations(), PLearn::RBMGaussianLayer::computeExpectations(), PLearn::RBMBinomialLayer::computeExpectations(), PLearn::RBMModule::computeFreeEnergyOfHidden(), PLearn::KLp0p1RBMModule::computeFreeEnergyOfHidden(), PLearn::RBMModule::computeFreeEnergyOfVisible(), PLearn::KLp0p1RBMModule::computeFreeEnergyOfVisible(), PLearn::SummationKernel::computeGramMatrix(), PLearn::RationalQuadraticARDKernel::computeGramMatrix(), PLearn::PLearnerDiagonalKernel::computeGramMatrix(), PLearn::KroneckerBaseKernel::computeGramMatrix(), PLearn::Kernel::computeGramMatrix(), PLearn::IIDNoiseKernel::computeGramMatrix(), PLearn::RationalQuadraticARDKernel::computeGramMatrixDerivIspAlpha(), PLearn::Matern1ARDKernel::computeGramMatrixDerivIspInputSigma(), PLearn::RationalQuadraticARDKernel::computeGramMatrixDerivIspSignalSigma(), PLearn::PLearnerDiagonalKernel::computeGramMatrixDerivIspSignalSigma(), PLearn::Matern1ARDKernel::computeGramMatrixDerivIspSignalSigma(), PLearn::IIDNoiseKernel::computeGramMatrixDerivKronecker(), PLearn::MemoryCachedKernel::computeGramMatrixNV(), PLearn::RBMModule::computeHiddenActivations(), PLearn::KLp0p1RBMModule::computeHiddenActivations(), PLearn::LinearRegressor::computeInformationCriteria(), PLearn::computeInverseStandardDeviationFromMeanAndSquareMean(), PLearn::LayerCostModule::computeKLdiv(), PLearn::ReconstructionWeightsKernel::computeLLEMatrix(), PLearn::computeLocalPrincipalComponents(), PLearn::GaussMix::computeLogLikelihood(), PLearn::DeepNonLocalManifoldParzen::computeManifoldParzenParameters(), PLearn::computeMeanAndCovar(), PLearn::WPLS::computeOutput(), PLearn::PLS::computeOutput(), PLearn::ManifoldParzen2::computeOutput(), PLearn::KernelRidgeRegressor::computeOutput(), PLearn::ModuleLearner::computeOutputAndCosts(), PLearn::PLearner::computeOutputCovMat(), PLearn::GaussianProcessRegressor::computeOutputCovMat(), PLearn::NatGradNNet::computeOutputs(), PLearn::mNNet::computeOutputs(), PLearn::ModuleLearner::computeOutputsAndCosts(), PLearn::Kernel::computePartialGramMatrix(), PLearn::computePrincipalComponents(), PLearn::RBMSparse1DMatrixConnection::computeProducts(), PLearn::RBMMixedConnection::computeProducts(), PLearn::RBMMatrixConnection::computeProducts(), PLearn::RBMDiagonalMatrixConnection::computeProducts(), PLearn::RBMConv2DConnection::computeProducts(), PLearn::computeRanks(), PLearn::LinearRegressor::computeResidualsVariance(), PLearn::LocalizedFeaturesLayerVariable::computeSubsets(), PLearn::Kernel::computeTestGramMatrix(), PLearn::PruningLinearRegressor::computeTRatio(), PLearn::RBMModule::computeVisibleActivations(), PLearn::KLp0p1RBMModule::computeVisibleActivations(), PLearn::SumVarianceOfLinearTransformedBernoullis::computeWsqnorm(), PLearn::constrainedLinearRegression(), PLearn::convolve(), PLearn::convolve2D(), PLearn::convolve2Dbackprop(), PLearn::correlation(), PLearn::cross_valid(), PLearn::Variable::defineGradientLocation(), PLearn::Variable::defineValueLocation(), PLearn::det(), PLearn::diagonalizedFactorsProduct(), PLearn::diagonalizedFactorsProductBprop(), PLearn::diagonalizedFactorsProductTranspose(), PLearn::diagonalizedFactorsProductTransposeBprop(), PLearn::diagonalizedFactorsTransposeProduct(), PLearn::diagonalizedFactorsTransposeProductBprop(), PLearn::diagonalizedFactorsTransposeProductTranspose(), PLearn::diagonalizedFactorsTransposeProductTransposeBprop(), PLearn::diffSquareMultiplyAcc(), PLearn::DirichletEstimatorMMoments(), PLearn::GhostScript::displayBlack(), PLearn::GhostScript::displayGray(), PLearn::GhostScript::displayRGB(), PLearn::DX_write_2D_data(), PLearn::DX_write_2D_data_for_grid(), PLearn::eigen_SymmMat(), PLearn::eigenSparseNonSymmMat(), PLearn::eigenSparseSymmMat(), PLearn::eigenVecOfSymmMat(), PLearn::DenoisingRecurrentNet::encode_artificialData(), PLearn::DenoisingRecurrentNet::encodeAndCreateSupervisedSequence(), PLearn::equals(), PLearn::Kernel::estimateHistograms(), PLearn::ReconstructionWeightsKernel::evaluate_i_j(), PLearn::ReconstructionWeightsKernel::evaluate_sum_k_i_k_j(), PLearn::PLMPI::exchangeColumnBlocks(), PLearn::externalProduct(), PLearn::externalProductAcc(), PLearn::externalProductDivUpdate(), PLearn::externalProductMultUpdate(), PLearn::externalProductScaleAcc(), PLearn::RBMLateralBinomialLayer::externalSymetricProductAcc(), PLearn::GaussianProcessNLLVariable::fbpropFragments(), PLearn::Gnuplot::featureplot(), PLearn::fill_one_hot(), PLearn::LiftStatsCollector::finalize(), PLearn::SubsamplingDBN::fineTuningStep(), PLearn::DeepBeliefNet::fineTuningStep(), PLearn::PseudolikelihoodRBM::forget(), PLearn::PLearnDiff::forget(), PLearn::LinearInductiveTransferClassifier::forget(), PLearn::GradNNetLayerModule::forget(), PLearn::WeightedLogGaussian::fprop(), PLearn::TreeDBNModule::fprop(), PLearn::TransposedDoubleProductVariable::fprop(), PLearn::TraceVariable::fprop(), PLearn::SumVarianceOfLinearTransformedBernoullis::fprop(), PLearn::SumEntropyOfCategoricals::fprop(), PLearn::SumEntropyOfBernoullis::fprop(), PLearn::SoftmaxNLLCostModule::fprop(), PLearn::SoftmaxModule::fprop(), PLearn::ShuntingNNetLayerModule::fprop(), PLearn::ScaleGradientModule::fprop(), PLearn::RBMWoodsLayer::fprop(), PLearn::RBMModule::fprop(), PLearn::RBMMixedLayer::fprop(), PLearn::RBMLayer::fprop(), PLearn::RBMLateralBinomialLayer::fprop(), PLearn::RBMBinomialLayer::fprop(), PLearn::ProjectionErrorVariable::fprop(), PLearn::ProcessInputCostModule::fprop(), PLearn::NonDiagVariable::fprop(), PLearn::NllSemisphericalGaussianVariable::fprop(), PLearn::NllGeneralGaussianVariable::fprop(), PLearn::NLLCostModule::fprop(), PLearn::NetworkModule::fprop(), PLearn::ModuleStackModule::fprop(), PLearn::MatrixModule::fprop(), PLearn::LogSoftSoftMaxVariable::fprop(), PLearn::LinearFilterModule::fprop(), PLearn::LayerCostModule::fprop(), PLearn::KLp0p1RBMModule::fprop(), PLearn::InsertZerosVariable::fprop(), PLearn::IdentityModule::fprop(), PLearn::GradNNetLayerModule::fprop(), PLearn::GaussianProcessNLLVariable::fprop(), PLearn::DoubleProductVariable::fprop(), PLearn::Cov2CorrVariable::fprop(), PLearn::ConstrainVariable::fprop(), PLearn::CombiningCostsModule::fprop(), PLearn::BinarizeModule::fprop(), PLearn::RBMWoodsLayer::fpropNLL(), PLearn::RBMMultinomialLayer::fpropNLL(), PLearn::RBMMixedLayer::fpropNLL(), PLearn::RBMLocalMultinomialLayer::fpropNLL(), PLearn::RBMLayer::fpropNLL(), PLearn::RBMLateralBinomialLayer::fpropNLL(), PLearn::RBMGaussianLayer::fpropNLL(), PLearn::RBMBinomialLayer::fpropNLL(), PLearn::SumOverBagsVariable::fpropOneBag(), PLearn::TreeDBNModule::full_fprop(), PLearn::GCV(), PLearn::generalizedEigenVecOfSymmMat(), PLearn::RBMDistribution::generate(), PLearn::RBMDistribution::generateN(), PLearn::PDistribution::generateN(), PLearn::RBMWoodsLayer::generateSamples(), PLearn::RBMTruncExpLayer::generateSamples(), PLearn::RBMRateLayer::generateSamples(), PLearn::RBMMultinomialLayer::generateSamples(), PLearn::RBMLocalMultinomialLayer::generateSamples(), PLearn::RBMLateralBinomialLayer::generateSamples(), PLearn::RBMGaussianLayer::generateSamples(), PLearn::RBMBinomialLayer::generateSamples(), PLearn::geometric_mean(), PLearn::VecStatsCollector::getCorrelation(), PLearn::VecStatsCollector::getCovariance(), PLearn::VMatrixFromDistribution::getMat(), PLearn::VMatrix::getMat(), PLearn::TransposeVMatrix::getMat(), PLearn::SubVMatrix::getMat(), PLearn::MemoryVMatrix::getMat(), PLearn::RemapLastColumnVMatrix::getNewRow(), PLearn::PairsVMatrix::getNewRow(), PLearn::AppendNeighborsVMatrix::getNewRow(), PLearn::MatrixModule::getPortSizes(), PLearn::OnlineGramNaturalGradientOptimizer::gramEigenNaturalGradient(), PLearn::grep(), PLearn::InferenceRBM::hiddenExpGivenInput(), PLearn::InferenceRBM::hiddenExpGivenInputTarget(), PLearn::InferenceRBM::hiddenExpGivenVisible(), PLearn::identityMatrix(), PLearn::NeighborhoodImputationVMatrix::impute(), PLearn::inverse(), PLearn::invertElements(), PLearn::TMat< T >::isEqual(), PLearn::jacobi(), PLearn::kernelPCAfromDotProducts(), PLearn::lapackCholeskyDecompositionInPlace(), PLearn::lapackCholeskySolveInPlace(), PLearn::lapackEIGEN(), PLearn::lapackGeneralizedEIGEN(), PLearn::lapackSolveLinearSystem(), PLearn::lapackSVD(), PLearn::layerBpropUpdate(), PLearn::layerL1BpropUpdate(), PLearn::layerL2BpropUpdate(), PLearn::leftPseudoInverse(), PLearn::linearRegression(), PLearn::linearRegressionNoBias(), PLearn::loadAsciiAsVMat(), PLearn::loadAsciiWithoutSize(), PLearn::loadATT800(), PLearn::loadBreastCancerWisconsin(), PLearn::loadClassificationDataset(), PLearn::loadCorel(), PLearn::loadCorelDatamat(), PLearn::loadDiabetes(), PLearn::loadLetters(), PLearn::loadPimaIndians(), PLearn::loadPMat(), PLearn::loadSTATLOG(), PLearn::loadUCIMLDB(), PLearn::loadUCISet(), PLearn::LU_decomposition(), main(), PLearn::makeItSymmetric(), PLearn::Variable::makeSharedGradient(), PLearn::Variable::makeSharedRValue(), PLearn::Variable::makeSharedValue(), PLearn::matInvert(), PLearn::matlabSave(), PLearn::matRowDotVec(), PLearn::matRowsDots(), PLearn::matRowsDotsAcc(), PLearn::MatTPlusSumSquaredVec< MatT >::MatTPlusSumSquaredVec(), PLearn::max(), PLearn::maxabs(), PLearn::mean(), PLearn::MemoryVMatrix::MemoryVMatrix(), PLearn::VecStatsCollector::merge(), PLearn::metricMultiDimensionalScaling(), PLearn::min(), PLearn::minabs(), PLearn::GaussMix::missingExpectation(), PLearn::Molecule::Molecule(), PLearn::multinomial_sample(), PLearn::multiply(), PLearn::multiplyAcc(), PLearn::multiplyColumns(), PLearn::multiplyScaledAdd(), PLearn::OnlineLearningModule::namedBpropAccUpdate(), PLearn::negateElements(), PLearn::newIndexedMatArray(), PLearn::ConvertToPyObject< Mat >::newPyObject(), PLearn::RBMTrainer::NLL(), PLearn::normalize(), PLearn::normalizeColumns(), PLearn::normalizeDataSet(), PLearn::normalizeDataSets(), PLearn::FractionSplitter::nSetsPerSplit(), PLearn::SubsamplingDBN::onlineStep(), PLearn::StackedAutoassociatorsNet::onlineStep(), PLearn::NatGradSMPNNet::onlineStep(), PLearn::NatGradNNet::onlineStep(), PLearn::DeepBeliefNet::onlineStep(), PLearn::operator*(), PLearn::operator*=(), PLearn::operator+(), PLearn::operator+=(), PLearn::operator-(), PLearn::operator-=(), PLearn::operator/=(), PLearn::operator<<(), PLearn::TMat< T >::operator==(), PLearn::operator^(), PLearn::RBMMatrixConnection::petiteCulotteOlivierCD(), PLearn::product(), PLearn::product2Acc(), PLearn::product2Transpose(), PLearn::product2TransposeAcc(), PLearn::productAcc(), PLearn::productScaleAcc(), PLearn::productTranspose(), PLearn::productTransposeAcc(), PLearn::productTransposeScaleAcc(), PLearn::VMatrix::putMat(), PLearn::SubVMatrix::putMat(), PLearn::MemoryVMatrix::putMat(), PLearn::qld_interface(), PLearn::RBMTrainer::recError(), PLearn::MatRowVariable::recomputeSize(), PLearn::regulargrid_x_y_outputs_to_bitmap(), PLearn::remote_solveLinearSystemByCholesky(), PLearn::removeColumn(), PLearn::LinearRegressor::resetAccumulators(), PLearn::ridgeRegressionByGCV(), PLearn::rightPseudoInverse(), PLearn::rowArgmax(), PLearn::rowArgmin(), PLearn::rowMax(), PLearn::rowMean(), PLearn::rowMin(), PLearn::rowSum(), PLearn::rowSumAcc(), PLearn::rowSumOfSquares(), PLearn::rowVariance(), PLearn::ExtractOptionCommand::run(), PLearn::ChemicalICP::run(), PLearn::saveAscii(), PLearn::saveAsciiWithoutSize(), PLearn::saveGnuplot(), PLearn::savePMat(), PLearn::saveSNMat(), PLearn::select(), PLearn::selectAndOrder(), PLearn::selectColumns(), PLearn::RBMConnection::setAsDownInputs(), PLearn::RBMConnection::setAsUpInputs(), PLearn::RBMLayer::setBatchSize(), PLearn::MatrixModule::setGradientTo(), PLearn::GaussMix::setPredictor(), PLearn::GaussMix::setPredictorPredictedSizes_const(), PLearn::TransformationLearner::setTransformsParameters(), PLearn::Variable::setValueSubMat(), PLearn::smooth(), PLearn::softmax(), PLearn::solveLinearSystemByCholesky(), PLearn::solveTransposeLinearSystemByCholesky(), PLearn::sortColumns(), PLearn::RegressionTreeRegisters::sortRows(), PLearn::SparseMatrix::SparseMatrix(), PLearn::DenoisingRecurrentNet::splitRawMaskedSupervisedSequence(), PLearn::sqrt(), PLearn::square(), PLearn::squareMultiplyAcc(), PLearn::squareProductAcc(), PLearn::squareProductTranspose(), PLearn::squareProductTransposeAcc(), PLearn::subsample(), PLearn::substract(), PLearn::substractFromColumns(), PLearn::sum(), PLearn::sum_of_squares(), PLearn::InferenceRBM::supCDStep(), PLearn::InferenceRBM::targetExpGivenInput(), PLearn::testSpearmanRankCorrelation(), PLearn::ThresholdedKernel::thresholdGramMatrix(), PLearn::TMat< T >::TMat(), PLearn::StructuralLearner::train(), PLearn::StackedSVDNet::train(), PLearn::PseudolikelihoodRBM::train(), PLearn::NatGradSMPNNet::train(), PLearn::NatGradNNet::train(), PLearn::mNNet::train(), PLearn::KernelRidgeRegressor::train(), PLearn::KernelProjection::train(), PLearn::GaussMix::train(), PLearn::GaussianDistribution::train(), PLearn::ModuleLearner::trainingStep(), PLearn::GaussianProcessRegressor::trainProjectedProcess(), PLearn::DenoisingRecurrentNet::trainUnconditionalPredictor(), PLearn::transpose(), PLearn::transposedLayerL1BpropUpdate(), PLearn::transposedLayerL2BpropUpdate(), PLearn::transposeProduct(), PLearn::transposeProduct2(), PLearn::transposeProduct2Acc(), PLearn::transposeProductAcc(), PLearn::transposeProductScaleAcc(), PLearn::transposeTransposeProduct(), PLearn::transposeTransposeProductAcc(), PLearn::transposeTransposeProductScaleAcc(), PLearn::RBMSparse1DMatrixConnection::update(), PLearn::RBMMatrixTransposeConnection::update(), PLearn::RBMMatrixConnectionNatGrad::update(), PLearn::RBMMatrixConnection::update(), PLearn::RBMLLParameters::update(), PLearn::RBMGaussianLayer::update(), PLearn::RBMDiagonalMatrixConnection::update(), PLearn::RBMConv2DConnection::update(), PLearn::NeuralProbabilisticLanguageModel::update_affine_transform(), PLearn::FeatureSetSequentialCRF::update_affine_transform(), PLearn::FeatureSetNNet::update_affine_transform(), PLearn::RBMLayer::updateCDandGibbs(), PLearn::ToBagClassifier::updateCostAndBagOutput(), PLearn::RBMMatrixConnection::updateGibbs(), PLearn::RBMLayer::updateGibbs(), PLearn::GaussMix::updateInverseVarianceFromPrevious(), PLearn::variance(), PLearn::NeuralProbabilisticLanguageModel::verify_gradient_affine_transform(), PLearn::FeatureSetSequentialCRF::verify_gradient_affine_transform(), PLearn::FeatureSetNNet::verify_gradient_affine_transform(), PLearn::weightedLinearRegression(), PLearn::weightedRidgeRegressionByGCV(), PLearn::GhostScript::writeBitmapHexString1Bit(), PLearn::GhostScript::writeBitmapHexString24Bits(), and PLearn::GhostScript::writeBitmapHexString8Bits().

    { return width_; }
template<class T>
void PLearn::TMat< T >::write ( PStream out) const [inline]

writes the Mat to the PStream: Note that users should rather use the form out << m;

Definition at line 370 of file TMat_decl.h.

    {
        T* ptr = 0;
        if(storage)
            ptr = data();

        switch(out.outmode)
        {
        case PStream::raw_ascii:      
        case PStream::pretty_ascii:
            for(int i=0; i<length_; i++, ptr+=mod_)
            {
                for(int j=0; j<width_; j++)
                {
                    out << ptr[j];
                    out.put('\t');
                }
                out.put('\n');
            }
            break;
        
        case PStream::raw_binary:
            for(int i=0; i<length_; i++, ptr+=mod_)
                binwrite_(out, ptr, width_);
            break;
        
        case PStream::plearn_ascii:
        {
            if(!out.implicit_storage)
            {
                out.write("TMat("); 
                out << length_ << width_ << mod_ << offset_ << storage;
                out.write(")\n");
            }
            else // implicit storage
            {
                out << length_;
                out.put(' ');
                out << width_;
                out.write(" [ \n");
                for(int i=0; i<length_; i++, ptr+=mod_)
                {
                    for(int j=0; j<width_; j++)
                    {
                        out << ptr[j];
                        out.put('\t');
                    }
                    out.put('\n');
                }
                out.write("]\n");
            }
        }
        break;

        case PStream::plearn_binary:
        {
            if(!out.implicit_storage)
            {
                out.write("TMat("); 
                out << length_ << width_ << mod_ << offset_ << storage;
                out.write(")\n");
            }
            else // implicit storage
            {
                unsigned char typecode;
                if(byte_order()==LITTLE_ENDIAN_ORDER)
                {
                    out.put(0x14); // 2D little-endian 
                    typecode = TypeTraits<T>::little_endian_typecode();
                }
                else
                {
                    out.put(0x15); // 2D big-endian
                    typecode = TypeTraits<T>::big_endian_typecode();
                }
              
                // write typecode
                out.put(typecode);
              
                // write length and width in raw_binary 
                out.write((char*)&length_, sizeof(length_));
                out.write((char*)&width_, sizeof(width_));
              
                // write the data
                for(int i=0; i<length_; i++, ptr+=mod_)
                    binwrite_(out, ptr, width_);
            }
        }
        break;
      
        default:
            PLERROR("In TMat::write(PStream& out)  unknown outmode!!!!!!!!!");
            break;
        }
    }

Friends And Related Function Documentation

template<class T>
friend class TVec< T > [friend]

Definition at line 66 of file TMat_decl.h.

template<class T>
friend class VarArray [friend]

for makeShared hack... (to be cleaned)

Definition at line 68 of file TMat_decl.h.

template<class T>
friend class Variable [friend]

for makeShared hack... (to be cleaned)

Definition at line 67 of file TMat_decl.h.


Member Data Documentation

template<class T>
int PLearn::TMat< T >::length_ [protected]
template<class T>
int PLearn::TMat< T >::mod_ [protected]
template<class T>
int PLearn::TMat< T >::offset_ [protected]
template<class T>
PP< Storage<T> > PLearn::TMat< T >::storage [protected]
template<class T>
int PLearn::TMat< T >::width_ [protected]

The documentation for this class was generated from the following files:
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines