ZOJ - 1008 Gnome Tetravex



ZOJ - 1008
Time Limit: 10000MS Memory Limit: 32768KB 64bit IO Format: %lld & %llu

 Status

Description

Hart is engaged in playing an interesting game, Gnome Tetravex, these days. In the game, at the beginning, the player is given n*n squares. Each square is divided into four triangles marked four numbers (range from 0 to 9). In a square, the triangles are the left triangle, the top triangle, the right triangle and the bottom triangle. For example, Fig. 1 shows the initial state of 2*2 squares.


Fig. 1 The initial state with 2*2 squares

The player is required to move the squares to the termination state. In the termination state, any two adjoining squares should make the adjacent triangle marked with the same number. Fig. 2 shows one of the termination states of the above example.


Fig. 2 One termination state of the above example

It seems the game is not so hard. But indeed, Hart is not accomplished in the game. He can finish the easiest game successfully. When facing with a more complex game, he can find no way out.

One day, when Hart was playing a very complex game, he cried out, "The computer is making a goose of me. It's impossible to solve it." To such a poor player, the best way to help him is to tell him whether the game could be solved. If he is told the game is unsolvable, he needn't waste so much time on it.


Input

The input file consists of several game cases. The first line of each game case contains one integer n, 0 <= n <= 5, indicating the size of the game.

The following n*n lines describe the marking number of these triangles. Each line consists of four integers, which in order represent the top triangle, the right triangle, the bottom triangle and the left triangle of one square.

After the last game case, the integer 0 indicates the termination of the input data set.


Output

You should make the decision whether the game case could be solved. For each game case, print the game number, a colon, and a white space, then display your judgment. If the game is solvable, print the string "Possible". Otherwise, please print "Impossible" to indicate that there's no way to solve the problem.

Print a blank line between each game case.

Note: Any unwanted blank lines or white spaces are unacceptable.


Sample Input

2
5 9 1 4
4 4 5 6
6 8 5 4
0 4 4 3
2
1 1 1 1
2 2 2 2
3 3 3 3
4 4 4 4
0


Output for the Sample Input

Game 1: Possible

Game 2: Impossible

Source

Asia 2001, Shanghai (Mainland China)

这道题的时间限制是10s,看起来真是非常nuiblity,然后动手做了一下,一开始WA,后来TLE,然后TLE,然后还是TLE......。隔了几天终于忍不住看题解,说是需要将相同的方块放在一起,文中还慢悠悠地补了一句,说是TAOCP中有提到这种优化,看来以后还是要多看书啊。

       在漫长的TLE岁月里,积累了一些优化的方法,我相信他们是有用的,比如讲方块按照四个方向上的数字归类好。这样查找的时候效率会比较高。

       后来用map重新写了一下,增进了我对map的了解。

       在使用map的时候,你只需要重载一个<操作符,可想而知,其实所有的判等,大于,小于都是依靠这个操作符实现的。

       所以当我把方块的结构体写成下面这个样子的时候,代码果断WA了。

struct grid{
	int up,right,down,left;
	void Read(){
		scanf("%d%d%d%d",&up,&right,&down,&left);
	}
	bool operator < (const grid& cmper)const{
		return up>cmper.up;
	}
} g[size*size];
此时题目考虑两个元素是否相同只看他的up。

       所以当我把方块的结构体写成下面这个样子的时候,代码果断TLE了。

struct grid{
	int up,right,down,left;
	void Read(){
		scanf("%d%d%d%d",&up,&right,&down,&left);
	}
	bool operator < (const grid& cmper)const{
		if (up!=cmper.up)
			return up < cmper.up;
		if (right!=cmper.right)
			return right < cmper.right;
		if (down!=cmper.down)
			return down < cmper.down;
		if (left!=cmper.left)
			return left < cmper.left;
	}
} g[size*size];
此时一次比较最多需要比较四次。

       我甚至一度想过散列,因为在这道题目中,次序对我毫无用处。最后果然散列了。

struct grid{
	int up,right,down,left;
	int sum;
	void Read(){
		scanf("%d%d%d%d",&up,&right,&down,&left);
		sum = ((up*10 + right)*10 + down)*10 + left;
	}
	bool operator < (const grid& cmper)const{
		return sum < cmper.sum;
	}
} g[size*size];
想想我也是蠢,其实一个gird也许只用一个数字就能存储了。比较的时候比较对应位上的数字就好了,不过现在懒得修改了。


最后是AC代码,其实除了那几个坑,这就是一道很简单的题目。

# include <cstdio>
# include <map>
using namespace std;

const int size = 5 + 3;
const int debug = 0;

struct grid{
	int up,right,down,left;
	int sum;
	void Read(){
		scanf("%d%d%d%d",&up,&right,&down,&left);
		sum = ((up*10 + right)*10 + down)*10 + left;
	}
	bool operator < (const grid& cmper)const{
		return sum < cmper.sum;
	}
} g[size*size];
int n;
map<grid,int> mp;

bool match(int no){
	int up = no-n,left = no-1;
	bool ret = true;
	if (up>0)
		ret = ret&&g[no].up==g[up].down;
	if (left%n!=0)
		ret = ret&&g[no].left==g[left].right;
    return ret;	
}
bool dfs(int no){
	if (no==n*n+1)
		return true;
	map<grid,int>::iterator it; 
	for (it=mp.begin();it!=mp.end();it++)
	{
		if (it->second<=0)
			continue;
		g[no] = it->first;
	    if (match(no)){
			it->second--;
			if (dfs(no+1))
				return true;
			it->second++;
		}
	}
	return false;
}
int main()
{
	
	int i,ncase = 0;
	while(scanf("%d",&n)&&n){
		mp.clear();
		if (ncase++) printf("\n");
		for (i=0;i<n*n;i++)
		{
			grid tmp;
			tmp.Read();
			mp[tmp]++;
		}
	    if (dfs(1))
			printf("Game %d: Possible\n",ncase);
		else 
			printf("Game %d: Impossible\n",ncase);
	} 
    return 0;
}


速度也只是比超时快3秒,有时间再优化。


     

  • 0
    点赞
  • 0
    收藏
    觉得还不错? 一键收藏
  • 0
    评论

“相关推荐”对你有帮助么?

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
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