AcWing 345. 牛站 Cow Relays

由于我太菜了,不会矩阵乘法,所以给同样不会矩阵乘法同学的福利

首先发现这题点很多边很少,实际上有用的点 \(<= 2 * T\)(因为每条边会触及两个点嘛)

所以我们可以把点的范围缩到 \(2 * T\)来,然后...

算法1 Bellman - Ford O(NT)

什么,限制边数?那不就是可爱的 \(BellmanFord\)吗?

看看复杂度,嗯嗯 \(10 ^ 8\) 海星,常数超小的我肯定不用吸氧的

#pragma GCC optimize(2)
#include <cstdio>
#include <iostream>
#include <algorithm>
#include <cmath>
#include <cstring>
using namespace std;
const int N = 205, M = 105;
struct Edge{
	int u, v, w;
}e[M];
int m, n, s, t, adj[N], dis[N], bDis[N], tot;
void inline read(int &x) {
    x = 0;
    char s = getchar();
    while(s > '9' || s < '0') s = getchar();
    while(s <= '9' && s >= '0') x = x * 10 + s - '0', s = getchar();
}
int inline get(int &x) {
	return lower_bound(adj + 1, adj + 1 + tot, x) - adj;
}

int inline bellmanFord(){
    memset(dis, 0x3f, sizeof dis);
    dis[s] = 0;
    for(register int i = 1; i <= n; i++){
        memcpy(bDis, dis, sizeof dis);
        memset(dis, 0x3f, sizeof dis);
        for(register int j = 1; j <= m; j++){
            dis[e[j].v] = min(dis[e[j].v], bDis[e[j].u] + e[j].w);
            dis[e[j].u] = min(dis[e[j].u], bDis[e[j].v] + e[j].w);
        }
    }
    return dis[t];
}



int main(){
	read(n); read(m); read(s); read(t);
	for (register int i = 1; i <= m; i++) {
		read(e[i].w); read(e[i].u); read(e[i].v);
		adj[++tot] = e[i].u;
		adj[++tot] = e[i].v;
	}
	sort(adj + 1, adj + 1 + tot);
	tot = unique(adj + 1, adj + 1 + tot) - adj - 1;
	for (register int i = 1; i <= m; i++) {
		e[i].u = get(e[i].u), e[i].v = get(e[i].v);
	}
	s = get(s), t = get(t);
	printf("%d\n", bellmanFord());
	return 0;
}

真香

算法2 倍增 + Floyd O(T ^ 3 * log_2N)

据说这题正解要用矩阵乘法,可我不会,咋办呢?

不如用倍增的思想,把\(N\)拆成二进制下的多个\(1\),我们把每个\('1'\)最短路搞出来,然后拼出来最终的最短路,先预处理:

\(d[i][j][l]\) 表示从 \(i\)\(j\) 恰好经过 \(2 ^ l\) 条边的最短路。

初始化 \(d[i][j][0] = w[i][j]\),剩下为正无穷(注意是恰好 \(N\) 条边,所以 \(d[i][i][0]\) 也是非法状态)

转移也很好想:

\(d[i][j][l] = min(d[i][k][l - 1] + d[k][j][l - 1])\),对于一个状态 \(d[i][j][l]\),枚举中间点 \(k\) 即可,所以预处理复杂度 \(O(T ^ 3 * log_2N)\)

接下来用二进制拼起来就行辣~,设 \(g[i]\) 为这前几部走完后,从 \(s\)\(i\) 的最短路, \(f[i]\) 为当前到 \(i\) 的最短路,与保卫王国的拼凑法思想差不多,即:

\(f[i] = min(g[j] + d[j][i][c])\)\(N\) 的二进制第 \(c\) 位为 \(1\)

#include <cstdio>
#include <iostream>
#include <algorithm>
#include <cmath>
#include <cstring>
using namespace std;
const int N = 205, M = 105;
struct Edge{
	int u, v, w;
}e[M];
int m, n, s, t, adj[N], tot, d[N][N][20], f[N], g[N];
int L;

int inline get(int x) {
	return lower_bound(adj + 1, adj + 1 + tot, x) - adj;
}
int main(){
	memset(d, 0x3f, sizeof d);
	scanf("%d%d%d%d", &n, &m, &s, &t);
	L = log2(n);
	for (int i = 1; i <= m; i++) {
		scanf("%d%d%d", &e[i].w, &e[i].u, &e[i].v);
		adj[++tot] = e[i].u;
		adj[++tot] = e[i].v;
	}
	sort(adj + 1, adj + 1 + tot);
	tot = unique(adj + 1, adj + 1 + tot) - adj - 1;
	for (int i = 1; i <= m; i++) {
		int u = get(e[i].u), v = get(e[i].v), w = e[i].w;
		d[u][v][0] = d[v][u][0] = min(d[u][v][0], w);
	}
	s = get(s), t = get(t);
	
	for (int c = 1; c <= L; c++) {
		for (int i = 1; i <= tot; i++) {
			for (int j = 1; j <= tot; j++) {
				for (int k = 1; k <= tot; k++) {
					d[i][j][c] = min(d[i][j][c], d[i][k][c - 1] + d[k][j][c - 1]);
				}
			}
		}
	}
	
	memset(g, 0x3f, sizeof g);
	g[s] = 0;
	for (int c = 0; c <= L; c++) {
		if(n >> c & 1) {
			memset(f, 0x3f, sizeof f);
			for (int i = 1; i <= tot; i++) 
				for (int j = 1; j <= tot; j++)
					f[i] = min(f[i], g[j] + d[j][i][c]);
			memcpy(g, f, sizeof g);
		}
	}
	printf("%d\n", f[t]);
	return 0;
}
posted @ 2019-11-10 20:06  DMoRanSky  阅读(212)  评论(0编辑  收藏  举报