KMeans.h

Go to the documentation of this file.

/*
 * 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) 1999-2008 Gunnar Raetsch
 * Written (W) 2007-2009 Soeren Sonnenburg
 * Copyright (C) 1999-2009 Fraunhofer Institute FIRST and Max-Planck-Society
 */

#ifndef _KMEANS_H__
#define _KMEANS_H__

#include <stdio.h>
#include <shogun/lib/common.h>
#include <shogun/io/SGIO.h>
#include <shogun/features/SimpleFeatures.h>
#include <shogun/distance/Distance.h>
#include <shogun/machine/DistanceMachine.h>

namespace shogun
{
class CDistanceMachine;

class CKMeans : public CDistanceMachine
{
    public:
        CKMeans();

        CKMeans(int32_t k, CDistance* d);
        virtual ~CKMeans();

        virtual inline EClassifierType get_classifier_type() { return CT_KMEANS; }

        virtual bool load(FILE* srcfile);

        virtual bool save(FILE* dstfile);

        inline void set_k(int32_t p_k)
        {
            ASSERT(p_k>0);
            this->k=p_k;
        }

        inline int32_t get_k()
        {
            return k;
        }

        inline void set_max_iter(int32_t iter)
        {
            ASSERT(iter>0);
            max_iter=iter;
        }

        inline float64_t get_max_iter()
        {
            return max_iter;
        }

        SGVector<float64_t> get_radiuses() { return R; }

        SGMatrix<float64_t> get_cluster_centers()
        {
            /* return empty matrix if no radiuses are there (not trained yet) */
            if (!R.vector)
                return SGMatrix<float64_t>();

            CSimpleFeatures<float64_t>* lhs=
                (CSimpleFeatures<float64_t>*)distance->get_lhs();
            SGMatrix<float64_t> centers=lhs->get_feature_matrix();
            SG_UNREF(lhs);
            return centers;
        }

        inline int32_t get_dimensions()
        {
            return dimensions;
        }

        inline virtual const char* get_name() const { return "KMeans"; }

    protected:
        void clustknb(bool use_old_mus, float64_t *mus_start);

        virtual bool train_machine(CFeatures* data=NULL);

        virtual void store_model_features();

    private:
        void init();

    protected:
        int32_t max_iter;

        int32_t k;

        int32_t dimensions;

        SGVector<float64_t> R;
        
    private:
        /* temporary variable for weighting over the train data */
        SGVector<float64_t> Weights;

        /* temp variable for cluster centers */
        SGMatrix<float64_t> mus;

};
}
#endif