/*
----
This file is NOT part of SECONDO.
Authors: Greg Hamerly and Jonathan Drake
Feedback: hamerly@cs.baylor.edu
See: http://cs.baylor.edu/~hamerly/software/kmeans.php
Copyright 2014
The Annulus K-means algorithm is based on Hamerly's algorithm, but also 
sortsthe centers by their norms (distances from the origin). Doing this 
allows searching the centers using the norm of the point to exclude
centers that cannot be close.
----

//paragraph [1] Title: [{\Large \bf \begin{center}] [\end{center}}]
//[TOC] [\tableofcontents]

[1] Declarations for the CompareKmeans algorithm


1 Declarations for the CompareKmeans algorithm

*/

#ifndef COMPARE_KMEANS_H
#define COMPARE_KMEANS_H

/* Authors: Greg Hamerly and Jonathan Drake
 * Feedback: hamerly@cs.baylor.edu
 * See: http://cs.baylor.edu/~hamerly/software/kmeans.php
 * Copyright 2014
 *
 * The CompareKmeans algorithm is an implementation of Phillip's Compare-Means
 * algorithm.
 */

#include "original_space_kmeans.h"

class CompareKmeans : public OriginalSpaceKmeans {
    public:
        CompareKmeans() : centersDist2div4(NULL) {}
        virtual ~CompareKmeans() { free(); }
        virtual void free();
        virtual void initialize(Dataset const *aX, unsigned short aK, 
        unsigned short *initialAssignment, int aNumThreads);
        virtual std::string getName() const { return "compare"; }
    
    private:
        virtual int runThread(int threadId, int maxIterations);
        void update_center_dists(int threadId);

        // A matrix (implemented in a single array) of center-center squared
        // distances, divided by 4. By dividing by 4 early, we save divisions
        // later during the comparison of this value.
        double *centersDist2div4;
};

#endif