我正在尝试使我现有的Prim算法实现,以保持与源代码的跟踪距离。因为prim的算法和Dijkstra的算法几乎相同。我不知道我错过了什么。
我知道问题出在哪里,但我想不出来。
这是我的代码,我如何修改它来打印从源到所有其他顶点的最短距离。最短距离存储在名为dist[]的数组中
代码:
package Graphs;
import java.util.ArrayList;
public class Prims {
static int no_of_vertices = 0;
public static void main(String[] args) {
int[][] graph = {{0, 2, 0, 6, 0},
{2, 0, 3, 8, 5},
{0, 3, 0, 0, 7},
{6, 8, 0, 0, 9},
{0, 5, 7, 9, 0},
};
no_of_vertices = graph.length;
int [][] result = new int [no_of_vertices][no_of_vertices];
boolean[] visited = new boolean[no_of_vertices];
int dist[] = new int[no_of_vertices];
for (int i = 0; i < no_of_vertices; i++)
for (int j = 0; j < no_of_vertices; j++) {
result[i][j]= 0;
if (graph[i][j] == 0) {
graph[i][j] = Integer.MAX_VALUE;
}
}
for (int i = 0; i < no_of_vertices; i++) {
visited[i] = false;
dist[i] = 0;
}
ArrayList<String> arr = new ArrayList<String>();
int min;
visited[0] = true;
int counter = 0;
while (counter < no_of_vertices - 1) {
min = 999;
for (int i = 0; i < no_of_vertices; i++) {
if (visited[i] == true) {
for (int j = 0; j < no_of_vertices; j++) {
if (!visited[j] && min > graph[i][j]) {
min = graph[i][j];
dist[i] += min; // <------ Problem here
visited[j] = true;
arr.add("Src :" + i + " Destination : " + j
+ " Weight : " + min);
}
}
}
}
counter++;
}
for (int i = 0; i < no_of_vertices; i++) {
System.out.println("Source : 0" + " Destination : " + i
+ " distance : " + dist[i]);
}
for (String str : arr) {
System.out.println(str);
}
}
}距离数组的计算中存在错误,因为它忘记添加从源到目标的任何中间节点的距离。
发布于 2015-04-28 05:19:11
for (int j = 0; j < no_of_vertices; j++) {
if (!visited[j] && min > graph[i][j]) {
min = graph[i][j];
dist[i] += min; // <------ Problem here当然,中间边不会被添加,因为您只添加了当前边。你可能想要这样的东西:
if (dist[i] + graph[i][j] < dist[j])
dist[j] = dist[i] + graph[i][j];并去掉min变量。
尽管你的算法在我看来并不正确。您应该在每个步骤中选择d[]最小的节点,并更新该节点的邻居,如我上面所写的,然后将其标记为已拾取,并且永远不再拾取它。
https://stackoverflow.com/questions/29905551
复制相似问题