标签:图的遍历 图的深度优先遍历 图的广度优先遍历 数据结构
图的遍历算法 有两种 :深度优先搜索遍历 和 广度 优先搜索遍历。深度优先搜索遍历类似与 树的 先序遍历。广度优先搜索遍历类似与树的层序遍历。只不过 图 可以有 不连通的 节点,所以 得 遍历 整个顶点数组。
深搜遍历 总是 先访问当前节点的邻接点,而 广搜算法 是 先访问顶点的邻接点 要 先于 后访问顶点的邻接点 被 访问。
具体遍历顺序如下:
以下代码 以 图的 邻接多重表 为 基本结构进行 遍历。
首先更改 上节 的 查找 邻接点 和 下一个邻接点的 返回值,以及 邻接点的 代码 有误,少加了 一句:
if (next->iIndex == location2 || next->jIndex == location2){
next = next->iIndex == location1 ? next->iNext : next->jNext;
break;
}
next = next->iIndex == location1 ? next->iNext : next->jNext;
int firstAdj(AMLGraph g,int location){ ArcNode * next = g.adjMuList[location].head->iNext; if (next != NULL) { int index = next->iIndex == location ? next->jIndex : next->iIndex; return index; } return -1; } int nextAdj(AMLGraph g,int location1 ,int location2){ ArcNode * next = g.adjMuList[location1].head->iNext; while (next != NULL){ if (next->iIndex == location2 || next->jIndex == location2){ next = next->iIndex == location1 ? next->iNext : next->jNext; break; } next = next->iIndex == location1 ? next->iNext : next->jNext; } if (next != NULL){ int index = next->iIndex == location1 ? next->jIndex : next->iIndex; return index; } return -1; }
void dfs(AMLGraph g,int i,bool * isVisitedArray){ printf("%c",g.adjMuList[i].vexName); isVisitedArray[i] = true; for (int next = firstAdj(g,i); next != -1 ; next = nextAdj(g,i,next)){ if (isVisitedArray[next] == false){ dfs(g,next,isVisitedArray); } } } //深度优先搜索遍历 void dfsTraver(AMLGraph g){ bool isVisited[MAX_VEX_NUM] = {false}; printf("----------深度优先遍历------------------\n"); for (int i = 0; i < g.vexNum; i++){ if (isVisited[i] == false){ dfs(g,i,isVisited); } } printf("\n"); } //广度优先搜索遍历 void bfsTraverse(AMLGraph g){ bool isVisited[MAX_VEX_NUM] = {false}; printf("----------广度优先遍历------------------\n"); LinkQueue queue; queueInit(&queue); for (int i = 0; i < g.vexNum; i++){ if (isVisited[i] == false){ printf("%c",g.adjMuList[i].vexName); isVisited[i] = true; enqueue(&queue,i); while (!queueEmpty(queue)){ int top; dequeue(&queue,&top); for (int next = firstAdj(g,top);next != -1 ; next = nextAdj(g,top,next)){ if (isVisited[next] == false){ printf("%c",g.adjMuList[next].vexName); isVisited[next] = true; enqueue(&queue,next); } } } } } queueDestory(&queue); }
广搜用到的链队代码没有放进来,想看的 可以 进网盘地址 下载 工程文件。
工程文件网盘地址:点击打开链接
// AMLGraph.cpp : 定义控制台应用程序的入口点。 //无向图的邻接多重表 #include "stdafx.h" #include <cstdlib> #include "queue.h" #define MAX_VEX_NUM 20 enum E_VisitIf { unvisited = 0, visited = 1, }; struct ArcNode { E_VisitIf mark; int iIndex,jIndex;//顶点i,j在图中的位置 ArcNode * iNext;//与i顶点点相关的下一个弧 ArcNode * jNext;//与j顶点点相关的下一个弧 }; struct VNode { char vexName; ArcNode * head;//头指针 }; struct AMLGraph { VNode adjMuList[MAX_VEX_NUM];//顶点数组 int vexNum,arcNum; }; //获取弧 的 头节点 ArcNode * getHeadNode(){ ArcNode * pNode = (ArcNode *)malloc(sizeof(ArcNode)); if (pNode){ pNode->iIndex = pNode->jIndex = -1; pNode->iNext = pNode->jNext = NULL; pNode->mark = unvisited; } return pNode; } ArcNode * getArcNode(int iIndex,int jIndex){ ArcNode * pNode = getHeadNode(); if (pNode){ pNode->iIndex = iIndex; pNode->jIndex = jIndex; } return pNode; } int vexLocation(AMLGraph g,char vex){ for (int i = 0; i < g.vexNum; i++){ if (g.adjMuList[i].vexName == vex){ return i; } } return -1; } void createGrahp(AMLGraph * g){ printf("输入图的顶点数 和 边(弧)数\n"); scanf("%d%d%*c",&g->vexNum,&g->arcNum); //构造顶点集 printf("请输入顶点集\n"); for (int i = 0; i < g->vexNum; i++){ char name; scanf("%c",&name); g->adjMuList[i].vexName = name; g->adjMuList[i].head = getHeadNode();//建立 头节点,并让头指针指向头节点 } //构造顶点关系 fflush(stdin); printf("请输入顶点的关系\n"); for (int i = 0; i < g->arcNum; i++){ char vex1,vex2; scanf("%c%c%*c",&vex1,&vex2); int location1 = vexLocation(*g,vex1); int location2 = vexLocation(*g,vex2); ArcNode * pNode = getArcNode(location1,location2); pNode->iNext = g->adjMuList[location1].head->iNext; g->adjMuList[location1].head->iNext = pNode; pNode->jNext = g->adjMuList[location2].head->iNext; g->adjMuList[location2].head->iNext = pNode; } } void destoryGraph(AMLGraph * g){ for (int i = 0; i < g->vexNum; i++){ ArcNode * next = g->adjMuList[i].head->iNext; while (next != NULL){ ArcNode * freeNode = next; next = next->iIndex == i ? next->iNext : next->jNext; if (freeNode->iIndex == i){////只释放 iIndex 等于 i的节点,要不会多次释放 free(freeNode); } } free(g->adjMuList[i].head); g->adjMuList[i].head = NULL; g->adjMuList[i].vexName = ' '; g->vexNum = g->arcNum = 0; } } //顶点vex1 和顶点vex2 是否相邻 bool graphIsAdj(AMLGraph g,char vex1,char vex2){ int location = vexLocation(g,vex1); ArcNode * next = g.adjMuList[location].head->iNext; while (next != NULL){ if (g.adjMuList[next->iIndex].vexName == vex2 || g.adjMuList[next->jIndex].vexName == vex2){ return true; } next = next->iIndex == location ? next->iNext : next->jNext; } return false; } int graphDegree(AMLGraph g,char vex){ int degree = 0; int location = vexLocation(g,vex); ArcNode * next = g.adjMuList[location].head->iNext;//计算所有出度 while (next != NULL){ degree++; next = next->iIndex == location ? next->iNext : next->jNext; } return degree; } //插入边(弧) void insertArc(AMLGraph * g,char vex1,char vex2){ int location1 = vexLocation(*g,vex1); int location2 = vexLocation(*g,vex2); ArcNode * node = getArcNode(location1,location2); node->iNext = g->adjMuList[location1].head->iNext; g->adjMuList[location1].head->iNext = node; node->jNext = g->adjMuList[location2].head->iNext; g->adjMuList[location2].head->iNext = node; g->arcNum ++; } //删除边(弧) void deleteArc(AMLGraph * g,char vex1,char vex2){ g->arcNum--; int location1 = vexLocation(*g,vex1); int location2 = vexLocation(*g,vex2); ArcNode * next = g->adjMuList[location1].head->iNext; ArcNode * pre = g->adjMuList[location1].head; while (next != NULL){ if (next->iIndex == location2){ if (pre == g->adjMuList[location1].head || pre->iIndex == location1){//删除的是第一个节点.或者 前驱的index = location1 pre->iNext = next->jNext; } else{ pre->jNext = next->jNext; } break; } else if(next->jIndex == location2){ if (pre == g->adjMuList[location1].head || pre->iIndex == location1){//删除的是第一个节点.或者 前驱的index = location1 pre->iNext = next->iNext; } else{ pre->jNext = next->iNext; } break; } pre = next; next = next->iIndex == location1 ? next->iNext : next->jNext; } next = g->adjMuList[location2].head->iNext; pre = g->adjMuList[location2].head; while (next != NULL){ if (next->iIndex == location1){ if (pre == g->adjMuList[location2].head || pre->iIndex == location2){//删除的是第一个节点.或者 前驱的index = location1 pre->iNext = next->jNext; } else{ pre->jNext = next->jNext; } free(next); break; } else if(next->jIndex == location1){ if (pre == g->adjMuList[location2].head || pre->iIndex == location2){//删除的是第一个节点.或者 前驱的index = location1 pre->iNext = next->iNext; } else{ pre->jNext = next->iNext; } free(next); break; } pre = next; next = next->iIndex == location2 ? next->iNext : next->jNext; } } //插入顶点 void insertVex(AMLGraph * g, char vex){ if (g->vexNum < MAX_VEX_NUM){ g->adjMuList[g->vexNum].vexName = vex; g->adjMuList[g->vexNum].head = getHeadNode(); g->vexNum++; } } //删除顶点 void deleteVex(AMLGraph * g,char vex){ int location = vexLocation(*g,vex); //删除顶点 同样需要 遍历整个 图 查找 与 vex 相关的弧节点 for (int i = 0; i < g->vexNum; i++){ ArcNode * next = g->adjMuList[i].head->iNext; while (next != NULL){ if (next->iIndex == location || next->jIndex == location){ ArcNode * delNode = next; next = next->iIndex == location ? next->iNext : next->jNext; char delData1 = g->adjMuList[delNode->iIndex].vexName; char delData2 = g->adjMuList[delNode->jIndex].vexName; deleteArc(g,delData1,delData2); } else{ next = next->iIndex == location ? next->iNext : next->jNext; } } } //更改因删除顶点 而导致的元素位置变化.. for (int i = 0; i < g->vexNum; i++){ ArcNode * next = g->adjMuList[i].head->iNext; while (next != NULL){ if (next->iIndex == i){ if(next->iIndex > location){ next->iIndex --; } if(next->jIndex > location){ next->jIndex --; } } next = next->iIndex == location ? next->iNext : next->jNext; } } free(g->adjMuList[location].head);//释放头节点 //vex下面的 顶点上移 for (int i = location + 1; i < g->vexNum; i++){ g->adjMuList[i-1] = g->adjMuList[i]; } g->vexNum --; } void printGrahp(AMLGraph g){ for (int i = 0; i < g.vexNum; i++){ printf("%c的 邻接点有:",g.adjMuList[i].vexName); ArcNode * next = g.adjMuList[i].head->iNext;//删除所有弧尾 while (next != NULL){ int index = next->iIndex == i ? next->jIndex : next->iIndex; printf("%c",g.adjMuList[index].vexName); next = next->iIndex == i ? next->iNext : next->jNext; } printf("\n"); } } int firstAdj(AMLGraph g,int location){ ArcNode * next = g.adjMuList[location].head->iNext; if (next != NULL) { int index = next->iIndex == location ? next->jIndex : next->iIndex; return index; } return -1; } int nextAdj(AMLGraph g,int location1 ,int location2){ ArcNode * next = g.adjMuList[location1].head->iNext; while (next != NULL){ if (next->iIndex == location2 || next->jIndex == location2){ next = next->iIndex == location1 ? next->iNext : next->jNext; break; } next = next->iIndex == location1 ? next->iNext : next->jNext; } if (next != NULL){ int index = next->iIndex == location1 ? next->jIndex : next->iIndex; return index; } return -1; } void dfs(AMLGraph g,int i,bool * isVisitedArray){ printf("%c",g.adjMuList[i].vexName); isVisitedArray[i] = true; for (int next = firstAdj(g,i); next != -1 ; next = nextAdj(g,i,next)){ if (isVisitedArray[next] == false){ dfs(g,next,isVisitedArray); } } } //深度优先搜索遍历 void dfsTraver(AMLGraph g){ bool isVisited[MAX_VEX_NUM] = {false}; printf("----------深度优先遍历------------------\n"); for (int i = 0; i < g.vexNum; i++){ if (isVisited[i] == false){ dfs(g,i,isVisited); } } printf("\n"); } //广度优先搜索遍历 void bfsTraverse(AMLGraph g){ bool isVisited[MAX_VEX_NUM] = {false}; printf("----------广度优先遍历------------------\n"); LinkQueue queue; queueInit(&queue); for (int i = 0; i < g.vexNum; i++){ if (isVisited[i] == false){ printf("%c",g.adjMuList[i].vexName); isVisited[i] = true; enqueue(&queue,i); while (!queueEmpty(queue)){ int top; dequeue(&queue,&top); for (int next = firstAdj(g,top);next != -1 ; next = nextAdj(g,top,next)){ if (isVisited[next] == false){ printf("%c",g.adjMuList[next].vexName); isVisited[next] = true; enqueue(&queue,next); } } } } } queueDestory(&queue); } //邻接多重表 int _tmain(int argc, _TCHAR* argv[]) { AMLGraph g; createGrahp(&g); printGrahp(g); dfsTraver(g); bfsTraverse(g); return 0; }运行截图:
从 a的邻接表 可以 看到 深搜的结果: 首先访问 a 节点,然后 访问 a的 第一个 邻接点d,然后 访问 d 的第一个邻接点c,然后 访问c的第一个邻接点e,然后访问 e的 邻接点
c和b ,c被访问过了,访问b,然后 访问 b的邻接点,等等.......最后 访问 单独的 顶点 fg.
所以 深搜结果为:adcebfg
广搜结果:先访问a,再访问a的所有邻接点dcb,访问d的所有邻接点ca(都被访问过了跳过),c的所有邻接点edba,只有e没被访问,访问他,然后访问b的所有邻接点:eca等等。。。最后访问单独的顶点 fg
所以广搜结果为:adcbefg
标签:图的遍历 图的深度优先遍历 图的广度优先遍历 数据结构
原文地址:http://blog.csdn.net/fuming0210sc/article/details/44976393