kruskal算法java实现_最小生成树-Kruskal算法 java代码实现

/*

*日期:2010-04-18 20:02

*开发者:heroyan

*联系方式:zndxysf@126.com

*功能:无向图最小生成树Kruskal算法实现案例

*/

import java.util.Scanner;

import java.util.Arrays;

import java.util.ArrayList;

public class Kruskal{

private static int MAX = 100;

private ArrayList edge = new ArrayList();//整个图的边

private ArrayList target = new ArrayList();//目标边，最小生成树

private int[] parent = new int[MAX];//标志所在的集合

private static double INFINITY = 99999999.99;//定义无穷大

private double mincost = 0.0;//最小成本

private int n;//结点个数

public Kruskal(){}

public static void main(String args[]){

Kruskal sp = new Kruskal();

sp.init();

sp.kruskal();

sp.print();

}

//初始化

public void init(){

Scanner scan = new Scanner(System.in);

int p,q;

double w;

System.out.println("spanning tree begin!Input the node number:");

n = scan.nextInt();

System.out.println("Input the graph(-1,-1,-1 to exit)");

while(true){

p = scan.nextInt();

q = scan.nextInt();

w = scan.nextDouble();

if(p < 0 || q < 0 || w < 0){

break;

}

Edge e = new Edge();

e.start = p;

e.end = q;

e.cost = w;

edge.add(e);

}

mincost = 0.0;

for(int i = 1; i <= n; ++i){

parent[i] = i;

}

}

//集合合并

public void union(int j, int k){

for(int i = 1; i <= n; ++i){

if(parent[i] == j){

parent[i] = k;

}

}

}

//prim算法主体

public void kruskal(){

//找剩下的n-2条边

int i = 0;

while(i < n-1 && edge.size() > 0){

//每次取一最小边，并删除

double min = INFINITY;

int tag = 0;

Edge tmp = null;

for(int j = 0; j < edge.size(); ++j){

Edge tt = edge.get(j);

if(tt.cost < min){

min = tt.cost;

tmp = tt;

}

}

int jj = parent[tmp.start];

int kk = parent[tmp.end];

//去掉环

if(jj != kk){

++i;

target.add(tmp);

mincost += tmp.cost;

union(jj,kk);

}

edge.remove(tmp);

}

if(i != n-1){

System.out.println("no spanning tree");

}

}

//打印结果

public void print(){

for(int i = 0; i < target.size(); ++i){

Edge e = target.get(i);

System.out.println("the " + (i+1) + "th edge:" + e.start + "---" + e.end);

}

}

}

class Edge

{

public int start;//始边

public int end;//终边

public double cost;//权重

}