算法基础10 —— BFS (迷宫问题 + 一维坐标的移动 + 炸弹人 + 蒜头君回家)

前置知识 - 队列

什么是队?
在这里插入图片描述
在这里插入图片描述

STL —— queue

特点:先进先出(First In First Out)

方法:

  • push(x):将x入队
  • pop():队首元素出队
  • front( ):获取队首元素
  • back( ):获取队尾元素
  • empty( ):判断queue是否为空
  • size( ):队列长度
#include<queue>
#include<iostream>

using namespace std;

int main()
{
    queue<int> que;//定义队列
    
    que.push(1);
    que.push(2);
    que.push(3);
    
    cout << que.front() << endl;//1
    que.pop();
    cout << que.front() << endl;//2
    que.pop(); 
    cout << que.front() << endl;//3
    que.pop();
    
    cout << que.empty() << endl;//1
    
    return 0;
}
迷宫问题

问题描述:给定一个地图,从起点开始搜索达到终点的最短步数

假设迷宫的空地用’.'表示
障碍物用‘#’表示
迷宫可表示为:

..#.
....
..#.
.#..
...#

迷宫问题DFS写法:

#include <iostream>

using namespace std;

const int N = 10;
int ans = 0x3f3f3f3f;//表示从起点到终点的最小步数
int n,m;//n表示地图的行数,m表示地图的列数
char a[N][N];//存储迷宫
int vis[N][N];//标记数组,vis[i][j]=0表示未被访问,vis[i][j]=1表示已被访问
int startx,starty,endx,endy;//起点和终点
int dir[4][2] = {{0,1},{1,0},{0,-1},{-1,0}};//方向数组

void dfs(int x,int y,int step)//x表示当前在x行,y表示当前在y列,当前的坐标为(x,y),目前的步数是step
{
    if (x == endx && y == endy)//走到了终点
    {
        if (step < ans)
            ans = step;//更新最小的步数
            
        return;//回溯
    }
    
    for (int i = 0;i < 4;i++)//未到终点,沿右下左上进行搜索
    {
        int new_x = x + dir[i][0];//new_x = x + dir[0][0] = x + 0 = x;new_y = y + dir[0][1] = y + 1 = y + 1
        int new_y = y + dir[i][1];//坐标由(x,y) ——> (x,y + 1),实现坐标右移一列
        
        if (new_x < 1 || new_x > n || new_y < 1 || new_y > m) continue;//剪枝:新的坐标超过数组边界
        
        if (a[new_x][new_y] == '.' && vis[new_x][new_y] == 0)//当前的地方是空地且未被访问
        {
            vis[new_x][new_y] = 1;//修改标记数组表示已经被访问
            dfs(new_x,new_y,step + 1);//深搜
            vis[new_x][new_y] = 0;//恢复现场
        }
    }
}

int main()
{
    cin >> n >> m;//行和列
    cin >> startx >> starty >> endx >> endy;//输入起点(1,1)和终点(4,3)
    
    for (int i = 1;i <= n;i++)
        for (int j = 1;j <= m;j++)
            cin >> a[i][j];
    
    vis[startx][starty] = 1;//从起点开始搜索达到终点的最短步数        
    dfs(startx,starty,0);//位于起点是第0步
    cout << ans;
    return 0;
}

输入数据

5 4
1 1 4 3
..#.
....
..#.
.#..
...#

迷宫问题BFS写法:
思路:
1.首先初始化一个队列,并将起点(1,1)入队,此时的步数为0
在这里插入图片描述

2.初始坐标(1,1),经过一步之后可以到达坐标(1,2)或者(2,1),将这两点入队,然后队首元素出队。此时队列如图
在这里插入图片描述

3.(1,2)再走一步之后可以到达坐标(2,2),将这点入队,然后队首元素出队。此时队列如图
在这里插入图片描述

以此类推,可得到最小步数… …
在这里插入图片描述

在这里插入图片描述
在这里插入图片描述

#include <queue>
#include <iostream>

using namespace std;

int n,m;
const int N = 10;
int startx,starty,endx,endy;
int a[N][N];//用来存储地图,a[i][j]=0表示空地,否则为障碍物
int vis[N][N];//标记数组,vis[i][j]=0表示该点未被访问,vis[i][j]=1表示已被访问
int dir[4][2] = {{0,1},{1,0},{0,-1},{-1,0}};//方向数组

struct node
{
    int x;//表示当前在坐标(x,y)处,步数为step
    int y;
    int step;
};

void bfs(int x,int y,int step)
{
    queue<node> q;//定义一个可以存储以上结构体的队列
    
    struct node t;//结构体
    t.x = x;
    t.y = y;
    t.step = step;
    
    q.push(t);//将起点入队(1,1,0)
    vis[x][y] = 1;//修改标记数组
    while (!q.empty())//队列非空
    {
        t = q.front();//将队首的信息取出
        if (t.x == endx && t.y == endy)//到达终点
        {
            cout << t.step;
            break;
        }
        
        for (int k = 0;k < 4;k++)//未到达终点,将(x,y)四个方向的点加入队列
        {
            int newx = t.x + dir[k][0];
            int newy = t.y + dir[k][1];
            if (vis[newx][newy] == 0 && a[newx][newy] == 0)//该点未被访问过且为空地
            {
                vis[newx][newy] = 1;//修改标记数组
                
                struct node new_node;
                
                new_node.x = newx;//设置新的
                new_node.y = newy;
                new_node.step = t.step + 1;
                q.push(new_node);//新的结构体入队
            }
        }
        
        q.pop();//四个方向扩展完毕之后将队首元素出队
    }
}

int main()
{
    cin >> n >> m;
    for (int i = 1;i <= n;i++)
        for (int j = 1;j <= m;j++)
            cin >> a[i][j];
    cin >> startx >> starty >> endx >> endy;
    bfs(startx,starty,0);
    return 0;
}

BFS测试样例

5 4
0 0 1 0
0 0 0 0
0 0 1 0
0 1 0 0
0 0 0 1
1 1 4 3

比较BFS与DFS
  • 标记数组、方向数组大部分情况下是都需要的
  • 如果题目要求找到最短路径,DFS即使找到了出口也不能提前提交答案,它必须找到所有能走到终点的路径,最后选择最短的,所以DFS适合小数据。在最短路径的问题上BFS更快一点,但是代码长度也会更大一点
  • 如果题目要求是否可以从A点到达B点(连通性问题),DFS也可以解决这类问题,但并不能保证是通过最短的路径到达的B点。反之,BFS则可以保证是通过最短的路径由A到达B
例题

例1 Acwing 1112 迷宫

例2 Acwing 844 走迷宫

(提示:例1可以使用DFS搜索答案。例2使用DFS会超时,最短路一般用BFS,每次向四周扩散,记录步数,DFS统计最短,要么数据很小要么特殊情况)

  • 1
    点赞
  • 0
    收藏
    觉得还不错? 一键收藏
  • 0
    评论
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值