Gaussian.cpp

浏览该文件的文档.

/*
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 3 of the License, or
 * (at your option) any later version.
 *
 * Written (W) 2011 Alesis Novik
 * Written (W) 2014 Parijat Mazumdar
 * Copyright (C) 2011 Berlin Institute of Technology and Max-Planck-Society
 */
#include <shogun/lib/config.h>

#ifdef HAVE_LAPACK

#include <shogun/distributions/Gaussian.h>
#include <shogun/mathematics/Math.h>
#include <shogun/base/Parameter.h>

using namespace shogun;

CGaussian::CGaussian() : CDistribution(), m_constant(0), m_d(), m_u(), m_mean(), m_cov_type(FULL)
{
    register_params();
}

CGaussian::CGaussian(const SGVector<float64_t> mean, SGMatrix<float64_t> cov, ECovType cov_type)
 : CDistribution()
{
    ASSERT(mean.vlen==cov.num_rows)
    ASSERT(cov.num_rows==cov.num_cols)
    m_d=SGVector<float64_t>();
    m_u=SGMatrix<float64_t>();
    m_cov_type=cov_type;

    m_mean=mean;

    if (cov.num_rows==1)
        m_cov_type=SPHERICAL;

    decompose_cov(cov);
    init();
    register_params();
}

void CGaussian::init()
{
    m_constant=CMath::log(2*M_PI)*m_mean.vlen;
    switch (m_cov_type)
    {
        case FULL:
        case DIAG:
            for (int32_t i=0; i<m_d.vlen; i++)
                m_constant+=CMath::log(m_d.vector[i]);
            break;
        case SPHERICAL:
            m_constant+=m_mean.vlen*CMath::log(m_d.vector[0]);
            break;
    }
}

CGaussian::~CGaussian()
{
}

bool CGaussian::train(CFeatures* data)
{
    // init features with data if necessary and assure type is correct
    if (data)
    {
        if (!data->has_property(FP_DOT))
                SG_ERROR("Specified features are not of type CDotFeatures\n")
        set_features(data);
    }

    CDotFeatures* dotdata=(CDotFeatures *) data;
    m_mean=dotdata->get_mean();
    SGMatrix<float64_t> cov=dotdata->get_cov();
    decompose_cov(cov);
    init();
    return true;
}

int32_t CGaussian::get_num_model_parameters()
{
    switch (m_cov_type)
    {
        case FULL:
            return m_u.num_rows*m_u.num_cols+m_d.vlen+m_mean.vlen;
        case DIAG:
            return m_d.vlen+m_mean.vlen;
        case SPHERICAL:
            return 1+m_mean.vlen;
    }
    return 0;
}

float64_t CGaussian::get_log_model_parameter(int32_t num_param)
{
    SG_NOTIMPLEMENTED
    return 0;
}

float64_t CGaussian::get_log_derivative(int32_t num_param, int32_t num_example)
{
    SG_NOTIMPLEMENTED
    return 0;
}

float64_t CGaussian::get_log_likelihood_example(int32_t num_example)
{
    ASSERT(features->has_property(FP_DOT))
    SGVector<float64_t> v=((CDotFeatures *)features)->get_computed_dot_feature_vector(num_example);
    float64_t answer=compute_log_PDF(v);
    return answer;
}

float64_t CGaussian::update_params_em(float64_t* alpha_k, int32_t len)
{
    CDotFeatures* dotdata=dynamic_cast<CDotFeatures *>(features);
    REQUIRE(dotdata,"dynamic cast from CFeatures to CDotFeatures returned NULL\n")
    int32_t num_dim=dotdata->get_dim_feature_space();

    // compute mean

    float64_t alpha_k_sum=0;
    SGVector<float64_t> mean(num_dim);
    mean.fill_vector(mean.vector,mean.vlen,0);
    for (int32_t i=0;i<len;i++)
    {
        alpha_k_sum+=alpha_k[i];
        SGVector<float64_t> v=dotdata->get_computed_dot_feature_vector(i);
        SGVector<float64_t>::add(mean.vector, alpha_k[i], v.vector, 1, mean.vector, v.vlen);
    }

    for (int32_t i=0; i<num_dim; i++)
        mean[i]/=alpha_k_sum;

    set_mean(mean);

    // compute covariance matrix

    float64_t* cov_sum=NULL;
    ECovType cov_type=get_cov_type();
    if (cov_type==FULL)
    {
        cov_sum=SG_MALLOC(float64_t, num_dim*num_dim);
        memset(cov_sum, 0, num_dim*num_dim*sizeof(float64_t));
    }
    else if(cov_type==DIAG)
    {
        cov_sum=SG_MALLOC(float64_t,num_dim);
        memset(cov_sum, 0, num_dim*sizeof(float64_t));
    }
    else if(cov_type==SPHERICAL)
    {
        cov_sum=SG_MALLOC(float64_t,1);
        cov_sum[0]=0;
    }

    for (int32_t j=0; j<len; j++)
    {
        SGVector<float64_t> v=dotdata->get_computed_dot_feature_vector(j);
        SGVector<float64_t>::add(v.vector, 1, v.vector, -1, mean.vector, v.vlen);

        switch (cov_type)
        {
            case FULL:
                cblas_dger(CblasRowMajor, num_dim, num_dim, alpha_k[j], v.vector, 1, v.vector,
                             1, (double*) cov_sum, num_dim);

                break;
            case DIAG:
                for (int32_t k=0; k<num_dim; k++)
                    cov_sum[k]+=v.vector[k]*v.vector[k]*alpha_k[j];

                break;
            case SPHERICAL:
                float64_t temp=0;

                for (int32_t k=0; k<num_dim; k++)
                    temp+=v.vector[k]*v.vector[k];

                cov_sum[0]+=temp*alpha_k[j];
                break;
        }
    }

    switch (cov_type)
    {
        case FULL:
            for (int32_t j=0; j<num_dim*num_dim; j++)
                cov_sum[j]/=alpha_k_sum;

            float64_t* d0;
            d0=SGMatrix<float64_t>::compute_eigenvectors(cov_sum, num_dim, num_dim);

            set_d(SGVector<float64_t>(d0, num_dim));
            set_u(SGMatrix<float64_t>(cov_sum, num_dim, num_dim));

            break;

        case DIAG:
            for (int32_t j=0; j<num_dim; j++)
                cov_sum[j]/=alpha_k_sum;

            set_d(SGVector<float64_t>(cov_sum,num_dim));

            break;

        case SPHERICAL:
            cov_sum[0]/=alpha_k_sum*num_dim;

            set_d(SGVector<float64_t>(cov_sum,1));

            break;
    }

    return alpha_k_sum;
}

float64_t CGaussian::compute_log_PDF(SGVector<float64_t> point)
{
    ASSERT(m_mean.vector && m_d.vector)
    ASSERT(point.vlen == m_mean.vlen)
    float64_t* difference=SG_MALLOC(float64_t, m_mean.vlen);
    memcpy(difference, point.vector, sizeof(float64_t)*m_mean.vlen);

    for (int32_t i = 0; i < m_mean.vlen; i++)
        difference[i] -= m_mean.vector[i];

    float64_t answer=m_constant;

    if (m_cov_type==FULL)
    {
        float64_t* temp_holder=SG_MALLOC(float64_t, m_d.vlen);
        cblas_dgemv(CblasRowMajor, CblasNoTrans, m_d.vlen, m_d.vlen,
                    1, m_u.matrix, m_d.vlen, difference, 1, 0, temp_holder, 1);

        for (int32_t i=0; i<m_d.vlen; i++)
            answer+=temp_holder[i]*temp_holder[i]/m_d.vector[i];

        SG_FREE(temp_holder);
    }
    else if (m_cov_type==DIAG)
    {
        for (int32_t i=0; i<m_mean.vlen; i++)
            answer+=difference[i]*difference[i]/m_d.vector[i];
    }
    else
    {
        for (int32_t i=0; i<m_mean.vlen; i++)
            answer+=difference[i]*difference[i]/m_d.vector[0];
    }

    SG_FREE(difference);

    return -0.5*answer;
}

SGVector<float64_t> CGaussian::get_mean()
{
    return m_mean;
}

void CGaussian::set_mean(SGVector<float64_t> mean)
{
    if (mean.vlen==1)
        m_cov_type=SPHERICAL;

    m_mean=mean;
}

void CGaussian::set_cov(SGMatrix<float64_t> cov)
{
    ASSERT(cov.num_rows==cov.num_cols)
    ASSERT(cov.num_rows==m_mean.vlen)
    decompose_cov(cov);
    init();
}

void CGaussian::set_d(const SGVector<float64_t> d)
{
    m_d = d;
    init();
}

SGMatrix<float64_t> CGaussian::get_cov()
{
    float64_t* cov=SG_MALLOC(float64_t, m_mean.vlen*m_mean.vlen);
    memset(cov, 0, sizeof(float64_t)*m_mean.vlen*m_mean.vlen);

    if (m_cov_type==FULL)
    {
        if (!m_u.matrix)
            SG_ERROR("Unitary matrix not set\n")

        float64_t* temp_holder=SG_MALLOC(float64_t, m_d.vlen*m_d.vlen);
        float64_t* diag_holder=SG_MALLOC(float64_t, m_d.vlen*m_d.vlen);
        memset(diag_holder, 0, sizeof(float64_t)*m_d.vlen*m_d.vlen);
        for(int32_t i=0; i<m_d.vlen; i++)
            diag_holder[i*m_d.vlen+i]=m_d.vector[i];

        cblas_dgemm(CblasRowMajor, CblasTrans, CblasNoTrans,
                    m_d.vlen, m_d.vlen, m_d.vlen, 1, m_u.matrix, m_d.vlen,
                    diag_holder, m_d.vlen, 0, temp_holder, m_d.vlen);
        cblas_dgemm(CblasRowMajor, CblasNoTrans, CblasNoTrans,
                    m_d.vlen, m_d.vlen, m_d.vlen, 1, temp_holder, m_d.vlen,
                    m_u.matrix, m_d.vlen, 0, cov, m_d.vlen);

        SG_FREE(diag_holder);
        SG_FREE(temp_holder);
    }
    else if (m_cov_type==DIAG)
    {
        for (int32_t i=0; i<m_d.vlen; i++)
            cov[i*m_d.vlen+i]=m_d.vector[i];
    }
    else
    {
        for (int32_t i=0; i<m_mean.vlen; i++)
            cov[i*m_mean.vlen+i]=m_d.vector[0];
    }
    return SGMatrix<float64_t>(cov, m_mean.vlen, m_mean.vlen, false);//fix needed
}

void CGaussian::register_params()
{
    SG_ADD(&m_u, "m_u", "Unitary matrix.",MS_NOT_AVAILABLE);
    SG_ADD(&m_d, "m_d", "Diagonal.",MS_NOT_AVAILABLE);
    SG_ADD(&m_mean, "m_mean", "Mean.",MS_NOT_AVAILABLE);
    SG_ADD(&m_constant, "m_constant", "Constant part.",MS_NOT_AVAILABLE);
    SG_ADD((machine_int_t*)&m_cov_type, "m_cov_type", "Covariance type.",MS_NOT_AVAILABLE);
}

void CGaussian::decompose_cov(SGMatrix<float64_t> cov)
{
    switch (m_cov_type)
    {
        case FULL:
            m_u=SGMatrix<float64_t>(cov.num_rows,cov.num_rows);
            memcpy(m_u.matrix, cov.matrix, sizeof(float64_t)*cov.num_rows*cov.num_rows);

            m_d.vector=SGMatrix<float64_t>::compute_eigenvectors(m_u.matrix, cov.num_rows, cov.num_rows);
            m_d.vlen=cov.num_rows;
            m_u.num_rows=cov.num_rows;
            m_u.num_cols=cov.num_rows;
            break;
        case DIAG:
            m_d=SGVector<float64_t>(cov.num_rows);
            for (int32_t i=0; i<cov.num_rows; i++)
                m_d[i]=cov.matrix[i*cov.num_rows+i];

            break;
        case SPHERICAL:
            m_d=SGVector<float64_t>(1);
            m_d.vector[0]=cov.matrix[0];
            break;
    }
}

SGVector<float64_t> CGaussian::sample()
{
    SG_DEBUG("Entering\n");
    float64_t* r_matrix=SG_MALLOC(float64_t, m_mean.vlen*m_mean.vlen);
    memset(r_matrix, 0, m_mean.vlen*m_mean.vlen*sizeof(float64_t));

    switch (m_cov_type)
    {
        case FULL:
        case DIAG:
            for (int32_t i=0; i<m_mean.vlen; i++)
                r_matrix[i*m_mean.vlen+i]=CMath::sqrt(m_d.vector[i]);

            break;
        case SPHERICAL:
            for (int32_t i=0; i<m_mean.vlen; i++)
                r_matrix[i*m_mean.vlen+i]=CMath::sqrt(m_d.vector[0]);

            break;
    }

    float64_t* random_vec=SG_MALLOC(float64_t, m_mean.vlen);

    for (int32_t i=0; i<m_mean.vlen; i++)
        random_vec[i]=CMath::randn_double();

    if (m_cov_type==FULL)
    {
        float64_t* temp_matrix=SG_MALLOC(float64_t, m_d.vlen*m_d.vlen);
        cblas_dgemm(CblasRowMajor, CblasNoTrans, CblasNoTrans,
                    m_d.vlen, m_d.vlen, m_d.vlen, 1, m_u.matrix, m_d.vlen,
                    r_matrix, m_d.vlen, 0, temp_matrix, m_d.vlen);
        SG_FREE(r_matrix);
        r_matrix=temp_matrix;
    }

    float64_t* samp=SG_MALLOC(float64_t, m_mean.vlen);

    cblas_dgemv(CblasRowMajor, CblasNoTrans, m_mean.vlen, m_mean.vlen,
                1, r_matrix, m_mean.vlen, random_vec, 1, 0, samp, 1);

    for (int32_t i=0; i<m_mean.vlen; i++)
        samp[i]+=m_mean.vector[i];

    SG_FREE(random_vec);
    SG_FREE(r_matrix);

    SG_DEBUG("Leaving\n");
    return SGVector<float64_t>(samp, m_mean.vlen, false);//fix needed
}

CGaussian* CGaussian::obtain_from_generic(CDistribution* distribution)
{
    if (!distribution)
        return NULL;

    CGaussian* casted=dynamic_cast<CGaussian*>(distribution);
    if (!casted)
        return NULL;

    /* since an additional reference is returned */
    SG_REF(casted);
    return casted;
}

#endif // HAVE_LAPACK