DiscreteKernel.h

Go to the documentation of this file.

//===========================================================================
/*!
 * 
 *
 * \brief       Kernel on a finite, discrete space.
 * 
 * 
 *
 * \author      T. Glasmachers
 * \date        2012
 *
 *
 * \par Copyright 1995-2017 Shark Development Team
 * 
 * <BR><HR>
 * This file is part of Shark.
 * <http://shark-ml.org/>
 * 
 * Shark is free software: you can redistribute it and/or modify
 * it under the terms of the GNU Lesser General Public License as published 
 * by the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 * 
 * Shark is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU Lesser General Public License for more details.
 * 
 * You should have received a copy of the GNU Lesser General Public License
 * along with Shark.  If not, see <http://www.gnu.org/licenses/>.
 *
 */
//===========================================================================

#ifndef SHARK_MODELS_KERNELS_DISCRETEKERNEL_H
#define SHARK_MODELS_KERNELS_DISCRETEKERNEL_H


#include <shark/Models/Kernels/AbstractKernelFunction.h>
#include <shark/LinAlg/Base.h>
#include <vector>

namespace shark {


///
/// \brief Kernel on a finite, discrete space.
///
/// \par
/// This class represents a kernel function on a finite
/// space with N elements. Wlog, this space is represented
/// by the integers 0, ..., N-1. The kernel function is
/// defined by a symmetric, positive semi-definite N x N
/// matrix.
///
class DiscreteKernel : public AbstractKernelFunction<std::size_t>
{
public:
    typedef AbstractKernelFunction<std::size_t> base_type;

    /// \brief Construction of the kernel from a positive definite, symmetric matrix.
    DiscreteKernel(RealMatrix const& matrix)
    : m_matrix(matrix)
    {
        SHARK_RUNTIME_CHECK(matrix.size1() == matrix.size2(), "[DiscreteKernel::DiscreteKernel] kernel matrix must be square");
#ifdef DEBUG
        for (std::size_t i=0; i<matrix.size1(); i++)
        {
            for (std::size_t j=0; j<i; j++)
            {
                SHARK_RUNTIME_CHECK(matrix(i, j) == matrix(j, i), "[DiscreteKernel::DiscreteKernel] kernel matrix must be symmetric");
            }
        }
#endif
    }

    /// \brief From INameable: return the class name.
    std::string name() const
    { return "DiscreteKernel"; }

    RealVector parameterVector() const
    {
        return RealVector();
    }

    void setParameterVector(RealVector const& newParameters)
    {
        SIZE_CHECK(newParameters.size() == 0);
    }

    std::size_t numberOfParameters() const
    {
        return 0;
    }

    /// \brief Cardinality of the discrete space.
    std::size_t size() const
    { return m_matrix.size1(); }
    
    ///\brief DiscreteKernels don't have a state so they return an EmptyState object
    boost::shared_ptr<State> createState()const{
        return boost::shared_ptr<State>(new EmptyState());
    }

    /// \brief Evaluates the kernel function.
    ///
    /// The function returns the stored similarity value.
    double eval(ConstInputReference x1, ConstInputReference x2)const
    {
        return m_matrix(x1, x2);
    }
    
    /// \brief Evaluates the kernel function for every point combination of the two batches
    ///
    /// The function returns the stored similarity value.
    void eval(ConstBatchInputReference batchX1, ConstBatchInputReference batchX2, RealMatrix& result, State& state) const{
        eval(batchX1,batchX2,result);
    }
    /// \brief Evaluates the kernel function for every point combination of the two batches
    ///
    /// The function returns the stored similarity value.
    void eval(ConstBatchInputReference batchX1, ConstBatchInputReference batchX2, RealMatrix& result) const{
        std::size_t sizeX1 = batchSize(batchX1);
        std::size_t sizeX2 = batchSize(batchX2);
        ensure_size(result,sizeX1,sizeX2);
        for(std::size_t i = 0; i != sizeX1; ++i)
            for(std::size_t j = 0; j != sizeX2; ++j)
                result(i,j)=m_matrix(i,j); 
    }


    /// From ISerializable.
    void read(InArchive& ar)
    { ar >> m_matrix; }

    /// From ISerializable.
    void write(OutArchive& ar) const
    { ar << m_matrix; }

protected:
    /// kernel matrix
    RealMatrix m_matrix;
};


}
#endif