邻接表实现--图的深度优先遍历DFS和广度优先遍历BFS
程序员文章站
2022-05-21 23:06:16
...
图论中一个基本的概念就是遍历。就是访问到图的每一个顶点,同时每个顶点只访问一次。
DFS和BFS的概念和思路网上说明的很详细了。但是网上很多代码实现有缺陷,基本都没有考虑图不连通的情况,比如某个顶点A和其它任何一个顶点都不关联,那么这个顶点A就访问不到了。如果遍历的起点刚好是孤立的顶点A,就只能访问顶点A了,其它顶点就访问不到了。
我这边的代码就是增加了这些情况的处理,确保每个顶点都能可以访问到。
完整的代码如下(通过邻接表实现图):
#define MAX_VERTEX 16
typedef enum
{
UNDIRECTED_GRAPH = 0, //无向图
DIRECTED_GRAPH = 1, //有向图
UNDIRECTED_NET = 2, //无向网
DIRECTED_NET = 3, //有向网
}GRAPH_TYPE;
typedef struct _ARC_NODE
{
int index; //邻接顶点在顶点数组中的索引值
struct _ARC_NODE* next;//指向下一个相邻顶点
}ARC_NODE,*PARC_NODE;
typedef struct _VERTEX
{
char data;//顶点数据值
PARC_NODE outHead;//出边表头指针
PARC_NODE inHead; //入边表头指针
}VERTEX,*PVERTEX;
typedef struct
{
GRAPH_TYPE type; //图的类型
int vertexCount;//顶点个数
BOOL visitFlag[MAX_VERTEX];
VERTEX vertex[MAX_VERTEX];
}LINKED_GRAPH,*PLINKED_GRAPH; //邻接表方式的图或者网
void DFS(PLINKED_GRAPH graph,int startVertexIndex);
void BFS(PLINKED_GRAPH graph,int startVertexIndex);
void Visit(PLINKED_GRAPH graph,int vIndex)
{
graph->visitFlag[vIndex] = TRUE; //设置已访问标志
printf("Visit Vertex: %c\r\n",graph->vertex[vIndex].data);
}
void InitVistFlag(PLINKED_GRAPH graph)
{
for(int i=0;i<graph->vertexCount;i++)
{
graph->visitFlag[i] = FALSE;
}
}
void DFS(PLINKED_GRAPH graph,int startVertexIndex)
{
stack<int> s; //访问栈
s.push(startVertexIndex);
while(!s.empty())
{
int vertexIndex = s.top();
if(!graph->visitFlag[vertexIndex])//未访问过
{
Visit(graph,vertexIndex);
}
s.pop();//vertexIndex顶点出栈
//这里我们通过出度表来遍历
PARC_NODE p = graph->vertex[vertexIndex].outHead;
while(p)
{
//与vertexIndex相邻的顶点并且未访问过的顶点全部入栈
if(!graph->visitFlag[p->index])
{
s.push(p->index);
}
p = p->next;//指向下一个与vertexIndex相邻的顶点
}
}
//图并不一定是连通的,因此要确保每个顶点都遍历过
for(int i=0;i<graph->vertexCount;i++)
{
if(!graph->visitFlag[i])
{
printf("Not Connected vertex start DFS: %c\r\n",graph->vertex[i]);
DFS(graph,i);
}
}
}
void BFS(PLINKED_GRAPH graph,int startVertexIndex)
{
queue<int> q; //访问队列
q.push(startVertexIndex);//起始访问的顶点入队
while(!q.empty())
{
int vertexIndex = q.front();
if(!graph->visitFlag[vertexIndex])//未访问过
{
Visit(graph,vertexIndex);
}
q.pop();//vertexIndex顶点出队
//这里我们通过出度表来遍历
PARC_NODE p = graph->vertex[vertexIndex].outHead;
while(p)
{
//与vertexIndex相邻的顶点并且未访问过的顶点全部入队
if(!graph->visitFlag[p->index])
{
q.push(p->index);
}
p = p->next;//指向下一个与vertexIndex相邻的顶点
}
}
//图并不一定是连通的,因此要确保每个顶点都遍历过
for(int i=0;i<graph->vertexCount;i++)
{
if(!graph->visitFlag[i])
{
printf("Not Connected vertex start BFS: %c\r\n",graph->vertex[i]);
BFS(graph,i);
}
}
}
void InitLinkedGraph(PLINKED_GRAPH graph)
{
graph->type = UNDIRECTED_GRAPH; //无向图
graph->vertexCount = 10;
for(int i=0;i<graph->vertexCount;i++)
{
graph->vertex[i].data = 'A'+i; //顶点为'A','B','C'等等
}
graph->vertex[0].outHead = new ARC_NODE;
graph->vertex[0].outHead->index = 1; //AB有边
graph->vertex[0].outHead->next = new ARC_NODE;
graph->vertex[0].outHead->next->index = 4;//AE有边
graph->vertex[0].outHead->next->next = NULL;
graph->vertex[1].outHead = new ARC_NODE;
graph->vertex[1].outHead->index = 0; //BA有边
graph->vertex[1].outHead->next = new ARC_NODE;
graph->vertex[1].outHead->next->index = 3;//BD有边
graph->vertex[1].outHead->next->next = NULL;
graph->vertex[2].outHead = new ARC_NODE;
graph->vertex[2].outHead->index = 4; //CE有边
graph->vertex[2].outHead->next = new ARC_NODE;
graph->vertex[2].outHead->next->index = 5;//CF有边
graph->vertex[2].outHead->next->next = new ARC_NODE;
graph->vertex[2].outHead->next->next->index = 6;//CG有边
graph->vertex[2].outHead->next->next->next = new ARC_NODE;
graph->vertex[2].outHead->next->next->next->index = 7; //CG有边
graph->vertex[2].outHead->next->next->next->next = NULL;
graph->vertex[3].outHead = new ARC_NODE;
graph->vertex[3].outHead->index = 1; //DB有边
graph->vertex[3].outHead->next = NULL;
graph->vertex[4].outHead = new ARC_NODE;
graph->vertex[4].outHead->index = 2; //EC有边
graph->vertex[4].outHead->next = NULL;
graph->vertex[5].outHead = new ARC_NODE;
graph->vertex[5].outHead->index = 2; //FC有边
graph->vertex[5].outHead->next = NULL;
graph->vertex[6].outHead = new ARC_NODE;
graph->vertex[6].outHead->index = 2; //GC有边
graph->vertex[6].outHead->next = new ARC_NODE;
graph->vertex[6].outHead->next->index = 8;//GI有边
graph->vertex[6].outHead->next->next = new ARC_NODE;
graph->vertex[6].outHead->next->next->index= 9;//GJ有边
graph->vertex[6].outHead->next->next->next = NULL;
graph->vertex[7].outHead = new ARC_NODE;
graph->vertex[7].outHead->index = 2; //HC有边
graph->vertex[7].outHead->next = NULL;
graph->vertex[8].outHead = new ARC_NODE;
graph->vertex[8].outHead->index = 6; //IG有边
graph->vertex[8].outHead->next = NULL;
graph->vertex[9].outHead = new ARC_NODE;
graph->vertex[9].outHead->index = 6; //JG有边
graph->vertex[9].outHead->next = NULL;
}
void PrintLinkedGraph(PLINKED_GRAPH graph)
{
printf("Linked Graph Info:\r\n");
for(int i=0;i<graph->vertexCount;i++)
{
printf("%2c",graph->vertex[i].data);
PARC_NODE p = graph->vertex[i].outHead;
while(p)
{
printf(" --> %2c",graph->vertex[p->index].data);
p = p->next;
}
printf("\r\n");
}
printf("\r\n");
}
void DestroyLinkedGraph(PLINKED_GRAPH graph)
{
for(int i=0;i<graph->vertexCount;i++)
{
PARC_NODE p = graph->vertex[i].outHead;
while(p)
{
PARC_NODE pNext = p->next;
p->next = NULL;
delete p;
p = pNext;
}
}
}
void TestLinkedGraph()
{
LINKED_GRAPH graph = {UNDIRECTED_GRAPH,0};
InitLinkedGraph(&graph);
InitVistFlag(&graph);
PrintLinkedGraph(&graph);
printf("Linked Graph DFS:\r\n");
DFS(&graph,0);
printf("\r\n");
InitVistFlag(&graph);
printf("Linked Graph DFS:\r\n");
BFS(&graph,0);
printf("\r\n");
DestroyLinkedGraph(&graph);
}
推荐阅读
-
Java实现二叉树的深度优先遍历和广度优先遍历算法示例
-
Java数据结构与算法:图、图的概念、深度优先搜索DFS、广度优先搜索BFS、思路分析、代码实现
-
【数据结构】图的遍历——深度优先搜索DFS、广度优先搜索BFS
-
数据结构与算法————图的遍历DFS深度优先搜索和BFS广度优先搜索
-
图的遍历---广度优先遍历和深度优先遍历
-
[PHP] 算法-邻接矩阵图的广度和深度优先遍历的PHP实现
-
python实现图的深度优先搜索(DFS)和广度优先搜索(BFS)算法及Dijkstra最短路径应用
-
实验3.4:以邻接表为存储结构的图的深度、宽度优先遍历
-
邻接表为存储结构 | 实现图的深度、宽度优先遍历
-
数据结构(图)——简单无向图的邻接矩阵,实现广度优先遍历