V 0.1

K-Means.c

#include <stdio.h>
#include <math.h>

const int data_total=6;
const int max_data=3;
/*double data_training[data_total][max_data]={
    {1,1,0,0},
    {0,0,0,1},
    {1,0,0,0},
    {0,0,1,1}
    //{2,3,1,0},
    //{1,9,0,1},
    //{0,1,2,1},
    //{9,1,1,5}
};*/
double data_training[data_total][max_data]={
    {1.1,1.7,1.8},
    {0,0,0},
    {0,0.5,1.5},
    {1,0,0},
    {0.5,0.5,0.5},
    {1,1,1}
};
const int cluster=2;
double centroid[cluster][max_data];

double euclidean_distance(int a, int b);

int main(){
    int initial_centroid[cluster];
    int cluster_group[data_total];

    for(int i=0; i<cluster; i++)
        scanf("%d", &initial_centroid[i]);

    for(int i=0; i<cluster; i++)
        for(int j=0; j<max_data; j++)
            centroid[i][j]=data_training[initial_centroid[i]-1][j];

    // for(int i=0; i<cluster; i++){
        // for(int j=0; j<max_data; j++)
            // printf("%lf ", centroid[i][j]);
        // printf("\n");
    // }

    
    for(int i=0; i<data_total; i++)
        cluster_group[i]=0;

    bool flag=true;
    //int xx=1;
    while(flag){
        // Save centroid for each cluster
        double tmp_centroid[cluster][max_data];
        for(int i=0; i<cluster; i++)
            for(int j=0; j<max_data; j++)
                tmp_centroid[i][j] = centroid[i][j];

        // Assigning all point to the closest centroid
        for(int i=0; i<data_total; i++){
            int tmp=0;
            double tmp2=9999999;
            for(int j=0; j<cluster; j++){
                double tmp3=euclidean_distance(i, j);
                // printf("tmp3 = %lf\n", tmp3);
                if(tmp2 > tmp3){
                    tmp2 = tmp3;
                    tmp = j+1;
                }
            }
            cluster_group[i] = tmp;
            // printf("tmp = %d\n", tmp);
        }

        // Recompute the centroid of each cluster
        for(int i=0; i<cluster; i++){
            for(int j=0; j<max_data; j++){
                int cnt = 0;
                double tmp = 0;
                for(int k=0; k<data_total; k++){
                    if(cluster_group[k]-1==i){
                        tmp += data_training[k][j];
                        cnt ++;
                    }
                }
                centroid[i][j] = tmp/cnt;
                // printf("%lf %d\n", tmp, cnt);
            }
        }

        // Check centroid
        bool flag_for = true;
        int i, j;
        for(i=0; i<cluster; i++){
            for(j=0; j<max_data; j++)
                if(tmp_centroid[i][j] != centroid[i][j]){
                    flag_for = false;
                    break;
                }
            if(!flag_for)
                break;
        }
        if(i==cluster && j==max_data)
            flag=false;
    }

    for(int i=0; i<data_total; i++)
        printf("Point %3d => Group %d\n", i,  cluster_group[i]);
    for(int i=0; i<cluster; i++){
        printf("Centroid ke-%d :", i+1);
        for(int j=0; j<max_data; j++)
            printf("%lf ", centroid[i][j]);
        printf("\n");
    }

    return 0;
}

double euclidean_distance(int a, int b){
    double total=0;
    for(int i=0; i<max_data; i++)
        total += pow((data_training[a][i]-centroid[b][i]), 2);
    return sqrt(total);
}