BallTree.cpp

Go to the documentation of this file.

/*
 * Copyright (c) The Shogun Machine Learning Toolbox
 * Written (w) 2014 Parijat Mazumdar
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *
 * 1. Redistributions of source code must retain the above copyright notice, this
 * list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright notice,
 * this list of conditions and the following disclaimer in the documentation
 * and/or other materials provided with the distribution.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
 * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
 * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 *
 * The views and conclusions contained in the software and documentation are those
 * of the authors and should not be interpreted as representing official policies,
 * either expressed or implied, of the Shogun Development Team.
 */

#include <shogun/multiclass/tree/BallTree.h>

using namespace shogun;

CBallTree::CBallTree(int32_t leaf_size, EDistanceType d)
: CNbodyTree(leaf_size,d)
{
}

float64_t CBallTree::min_dist(bnode_t* node,float64_t* feat, int32_t dim)
{
    float64_t dist=0;
    SGVector<float64_t> center=node->data.center;
    for (int32_t i=0;i<dim;i++)
        dist+=add_dim_dist(center[i]-feat[i]);

    dist=actual_dists(dist);
    return CMath::max(0.0,dist-node->data.radius);
}

float64_t CBallTree::min_dist_dual(bnode_t* nodeq, bnode_t* noder)
{
    float64_t dist=0;
    SGVector<float64_t> center1=nodeq->data.center;
    SGVector<float64_t> center2=noder->data.center;
    for (int32_t i=0;i<center1.vlen;i++)
        dist+=add_dim_dist(center1[i]-center2[i]);

    dist=actual_dists(dist);
    return CMath::max(0.0,dist-nodeq->data.radius-noder->data.radius);
}

float64_t CBallTree::max_dist_dual(bnode_t* nodeq, bnode_t* noder)
{
    float64_t dist=0;
    SGVector<float64_t> center1=nodeq->data.center;
    SGVector<float64_t> center2=noder->data.center;
    for (int32_t i=0;i<center1.vlen;i++)
        dist+=add_dim_dist(center1[i]-center2[i]);

    dist=actual_dists(dist);
    return (dist+nodeq->data.radius+noder->data.radius);
}

void CBallTree::min_max_dist(float64_t* pt, bnode_t* node, float64_t &lower,float64_t &upper, int32_t dim)
{
    float64_t dist=0;
    SGVector<float64_t> center=node->data.center;
    for (int32_t i=0;i<dim;i++)
        dist+=add_dim_dist(center[i]-pt[i]);

    dist=actual_dists(dist);
    lower=CMath::max(0.0,dist-node->data.radius);
    upper=dist+node->data.radius;
}

void CBallTree::init_node(bnode_t* node, index_t start, index_t end)
{
    SGVector<float64_t> upper_bounds(m_data.num_rows);
    SGVector<float64_t> lower_bounds(m_data.num_rows);

    SGVector<float64_t> center(m_data.num_rows);
    for (int32_t i=0;i<m_data.num_rows;i++)
    {
        center[i]=m_data(i,m_vec_id[start]);
        upper_bounds[i]=m_data(i,m_vec_id[start]);
        lower_bounds[i]=m_data(i,m_vec_id[start]);
        for (int32_t j=start+1;j<=end;j++)
        {
            float64_t data_pt=m_data(i,m_vec_id[j]);
            upper_bounds[i]=CMath::max(upper_bounds[i],data_pt);
            lower_bounds[i]=CMath::min(lower_bounds[i],data_pt);
            center[i]+=data_pt;
        }

        center[i]/=(end-start+1.f);
    }

    float64_t radius=0;
    for (int32_t i=start;i<=end;i++)
        radius=CMath::max(distance(m_vec_id[i],center.vector,center.vlen),radius);

    node->data.radius=radius;
    node->data.center=center;
    node->data.start_idx=start;
    node->data.bbox_upper=upper_bounds;
    node->data.bbox_lower=lower_bounds;
    node->data.end_idx=end;
}