图--遍历--广度/深度优先搜索

记录贴膜小哥的学习笔记和代码。

前言

图的遍历有广度优先搜索和深度优先搜索。这里用一个迷宫游戏作为例子。
五乘五的方格是一个迷宫，四周和白色方块为墙，灰色为路径。0为起点，24为终点。
这个迷宫就是一个简单的图，通过遍历解决问题。

广度优先搜索

广度优先搜索是图的层次遍历。他有如下特点：
1–占用空间大
2–不进行回溯
3–常常使用队列实现

完整代码

1–先求出迷宫的邻接矩阵。25 * 25的数组
每一行为一个节点，每一列为与其他节点的关系，1为有连接
2–用队列存储待遍历的节点
3–用链表存储已遍历的节点
4–初始化起点和终点
5–遍历队列，查找节点连接的节点，判断是否遍历过，放入队列

#include<iostream>
#include<queue>
#include<list>
#include<algorithm>
bool data[25][25] = {
	{ 0,1,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },//0
	{ 1,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,1,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 1,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },//5
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,1,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,1,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,0 },//10
	{ 0,0,0,0,0,0,0,0,0,0,1,0,1,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,1,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,1,0,0,0,0 },//15
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,1 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0 },//20
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0 },//24
};
typedef struct str_data
{
	int point;
	std::string path;
}STR_Data;
int m_end;
std::queue<STR_Data> m_queue;
std::list<int> m_list;
//start:开始位置
//end:结束位置
void BFS_init(int start,int end)
{
	char start_buf[10];
	sprintf_s(start_buf,"%d",start);
	STR_Data buff = {start,start_buf};
	if (m_queue.empty())
	{
		m_queue.push(buff);//初始位置
	}
	m_end = end;
}
STR_Data  BFS_run()
{
	while (!m_queue.empty())
	{
		STR_Data buf = m_queue.front();
		m_queue.pop();
		if (buf.point == m_end)//匹配成功
		{
			printf("ok\n");
			return  buf;
		}
		else
		{
			m_list.push_back(buf.point);//标记遍历过的节点
			//for (int i = buf.point + 1; i < 25; i++)
			for (int i = 0; i < 25; i++)
			{
				if (i == buf.point)
					continue;
				if (data[buf.point][i] == 1)//有连接
				{

					auto t= std::find(m_list.cbegin(), m_list.cend(),i);
					if (t == m_list.cend())
					{
						STR_Data str_buf;
						char s[10];
						str_buf.point = i;
						sprintf_s(s,"%d",i);
						str_buf.path = buf.path + "->"+s;
						m_queue.push(str_buf);
					}
				}
			}
		}
	}
	return {0,"0"};
}


int main()
{
	BFS_init(0,24);
	STR_Data buf=BFS_run();
	printf("%s\n", buf.path.c_str());
	system("pause");
	return 0;
}

深度优先搜索

深度优先搜索一直往下遍历，直到后面没有没遍历的节点后，回溯到上一个节点进行遍历。他有如下特点：
1–占用空间小
2–进行回溯
3–常常使用栈实现

完整代码

1–先求出迷宫的邻接矩阵。25 * 25的数组
2–用栈存储待遍历的节点
3–用链表存储已遍历的节点
4–初始化起点和终点
5–判断栈顶是否为终点，不是则循环递归，直到找到终点或者弹空。

#include<iostream>
#include<algorithm>
#include<list>
#include<stack>
bool data[25][25] = {
	{ 0,1,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },//0
	{ 1,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,1,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 1,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },//5
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,1,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,1,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,0 },//10
	{ 0,0,0,0,0,0,0,0,0,0,1,0,1,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,1,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,1,0,0,0,0 },//15
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,1 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0 },//20
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 },
	{ 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0 },//24
};

std::stack<int> m_stack;
std::list<int> m_list;
int m_end;
void DFS_init(int start ,int end)
{
	if (m_stack.empty())
	{
		m_stack.push(start);
	}
	m_end = end;
}

int DFS_run()
{
	if (m_stack.empty())
		return 0;
	int buf = m_stack.top();
	if (buf == m_end)
	{
		char re[200] = {0};
		printf("ok\n");
		m_stack.pop();
		sprintf_s(re, "%d",buf);
		while (!m_stack.empty())
		{
			int b = m_stack.top();
			char re2[200] = { 0 };
			memcpy(re2,re,200);
			sprintf_s(re,"%d -> %s",b,re2);
			m_stack.pop();
		}
		printf("%s\n",re);
		return 1;
	}
	else
	{
		m_list.push_back(buf);
		for (int i = 0; i < 25; i++)
		{
			if (i == buf)
				continue;
			if (data[buf][i] == 1)//有连接
			{
				auto t = std::find(m_list.cbegin(), m_list.cend(), i);
				if (t == m_list.cend())//新的节点
				{
					m_stack.push(i);
					if (DFS_run() == 1)
						return 1;
				}
			}
		}
		if(!m_stack.empty())
			m_stack.pop();
		return 0;
	}
}
int main()
{
	 DFS_init(0, 24);
	 DFS_run();
	 system("pause");
	 return 0;
}

其他

广度优先搜索的一个优点是，搜索到的终点是最优解，也就是最短距离。而深度优先搜索随机找一个路径进行搜索，不一定是最优解。如迷宫图，0->11有两条路。

广度优先搜索的结果为最优路径

深度优先搜索的结果为长路径