POJ 3162 Walking Race

Walking Race
Time Limit: 10000MS   Memory Limit: 131072K
Total Submissions: 4139   Accepted: 1037
Case Time Limit: 3000MS

Description

flymouse’s sister wc is very capable at sports and her favorite event is walking race. Chasing after the championship in an important competition, she comes to a training center to attend a training course. The center has N check-points numbered 1 through N. Some pairs of check-points are directly connected by two-way paths. The check-points and the paths form exactly a tree-like structure. The course lasts N days. On the i-th day, wc picks check-point i as the starting point and chooses another check-point as the finishing point and walks along the only simple path between the two points for the day’s training. Her choice of finishing point will make it that the resulting path will be the longest among those of all possible choices.

After every day’s training, flymouse will do a physical examination from which data will obtained and analyzed to help wc’s future training be better instructed. In order to make the results reliable, flymouse is not using data all from N days for analysis. flymouse’s model for analysis requires data from a series of consecutive days during which the difference between the longest and the shortest distances wc walks cannot exceed a bound M. The longer the series is, the more accurate the results are. flymouse wants to know the number of days in such a longest series. Can you do the job for him?

Input

The input contains a single test case. The test case starts with a line containing the integers N (N ≤ 106) and M (M < 109). Then follow N − 1 lines, each containing two integers fi and di (i = 1, 2, …, N− 1), meaning the check-points i + 1 and fi are connected by a path of length di.

Output

Output one line with only the desired number of days in the longest series.

Sample Input

3 2
1 1
1 3

Sample Output

3

Hint

Explanation for the sample:

There are three check-points. Two paths of lengths 1 and 3 connect check-points 2 and 3 to check-point 1. The three paths along with wc walks are 1-3, 2-1-3 and 3-1-2. And their lengths are 3, 4 and 4. Therefore data from all three days can be used for analysis.

Source

POJ Monthly--2006.12.31, flymouse

题意:一棵树,每条边有边权。首先需要求每个节点最远可以到达的距离,其次按节点的序号对这些距离排序。问你满足区间最大值减去最小值不大于m的连续区间最大是多少。
首先不难看出第一步求每个节点最远可以到达的距离,用树dp,XJB写下就可以了,那么问题就只有如何找到符合的区间,一开始这里卡了一下,后面发现其实只要初始l = r = 1,然后每次询问如果符合则r++,否则l++,r++,即可。这个也就需要n次询问所以说复杂度是完全够的。同时由于只是进行区间询问操作的,本想RMQ搞搞,结果MLE了,尴尬,老老实实谢了线段树。

//******************************************************************************************
//AC代码如下
//Ninaye
//******************************************************************************************
#include 
#include 
#include 
#include 
using namespace std;
typedef long long LL;
const LL INF = 1e18;
const int maxn = 1e6+10;
int n;
LL value[maxn],m;
struct DP{
	LL f_val,s_val_1,s_val_2;
	int son1,son2;
}dp[maxn];

int head[maxn],_cnt;
struct Edge{
	int to,next;
	LL w;
}edge[maxn<<1];
inline void init(){
	for(int i = 0; i <= n; i++){
		head[i] = -1;
		dp[i].son1 = dp[i].son2 = -1;
		dp[i].f_val = dp[i].s_val_1 = dp[i].s_val_2 = 0;
		value[i] = 0;
	}
	_cnt = 0;
}
inline void ADD(int u,int v,LL w){
	edge[_cnt].to = v;
	edge[_cnt].next = head[u];
	edge[_cnt].w = w;
	head[u] = _cnt++;
}
inline void dfs_1(int now,int fa){
	for(int i = head[now]; ~i; i = edge[i].next){
		int v = edge[i].to;
		if(v != fa){
			value[v] = edge[i].w;
			dfs_1(v,now);
			if(dp[now].son1 == -1){
				dp[now].son1 = v;
				dp[now].s_val_1 = dp[v].s_val_1+value[v];
			}
			else if(dp[now].son2 == -1){
				dp[now].son2 = v;
				dp[now].s_val_2 = dp[v].s_val_1+value[v];
				if(dp[now].s_val_1 < dp[now].s_val_2){
					swap(dp[now].s_val_1,dp[now].s_val_2);
					swap(dp[now].son1,dp[now].son2);
				}
			}
			else if(dp[now].s_val_2 < dp[v].s_val_1+value[v]){
				dp[now].son2 = v;
				dp[now].s_val_2 = dp[v].s_val_1+value[v];
				if(dp[now].s_val_1 < dp[now].s_val_2){
					swap(dp[now].s_val_1,dp[now].s_val_2);
					swap(dp[now].son1,dp[now].son2);
				}
			}
		}
	}
}
inline void dfs_2(int now,int fa){
	if(fa == 0)
		dp[now].f_val = 0;
	else if(now != dp[fa].son1)
		dp[now].f_val = max(dp[fa].s_val_1,dp[fa].f_val)+value[now];
	else
		dp[now].f_val = max(dp[fa].s_val_2,dp[fa].f_val)+value[now];
	for(int i = head[now]; ~i; i = edge[i].next){
		int v = edge[i].to;
		if(v != fa)
			dfs_2(v,now);
	}
}

struct Info{
	LL Min,Max;
	int l,r;
};
struct SGT{
	Info tree[maxn<<2];
	inline void push_up(int now){
		tree[now].Min = min(tree[now<<1].Min,tree[now<<1|1].Min);
		tree[now].Max = max(tree[now<<1].Max,tree[now<<1|1].Max);
	}
	inline void build(int now,int l,int r){
		tree[now].l = l;
		tree[now].r = r;
		if(l == r){
			tree[now].Min = tree[now].Max = value[l];
			return;
		}
		int Mid = (l+r)>>1;
		build(now<<1,l,Mid);
		build(now<<1|1,Mid+1,r);
		push_up(now);
	}
	inline LL que_Max(int now,int l,int r){
		if(l <= tree[now].l && tree[now].r <= r){
			return tree[now].Max;
		}
		LL ans = 0;
		int Mid = (tree[now].l + tree[now].r)>>1;
		if(l <= Mid)
			ans = max(ans,que_Max(now<<1,l,r));
		if(Mid < r)
			ans = max(ans,que_Max(now<<1|1,l,r));
		return ans;
	}
	inline LL que_Min(int now,int l,int r){
		if(l <= tree[now].l && tree[now].r <= r){
			return tree[now].Min;
		}
		LL ans = INF;
		int Mid = (tree[now].l + tree[now].r)>>1;
		if(l <= Mid)
			ans = min(ans,que_Min(now<<1,l,r));
		if(Mid < r)
			ans = min(ans,que_Min(now<<1|1,l,r));
		return ans;
	}
	inline LL que(int l,int r){
		return que_Max(1,l,r)-que_Min(1,l,r);
	}
}sgt;
int main(){
	int v;
	LL w;
	while(~scanf("%d %lld",&n,&m)){
		init();
		for(int i = 1; i < n; i++){
			scanf("%d %lld",&v,&w);
			ADD(i+1,v,w);
			ADD(v,i+1,w);
		}
		dfs_1(1,0);
		dfs_2(1,0);
		for(int i = 1; i <= n; i++)
			value[i] = max(dp[i].f_val,dp[i].s_val_1);
		sgt.build(1,1,n);
		int l,r,ans;
		ans = 0;
		l = r = 1;
		while(l <= r && r <= n){
			if(sgt.que(l,r) <= m){
				ans = r-l+1;
				r++;
			}
			else{
				l++,r++;
			}
		}
		printf("%d\n",ans);
	}
	return 0;
}
/*
input
11 2
1 1
2 2
2 3
4 1
5 2
1 1
7 2
8 3
9 1
9 2

output
3
*/



你可能感兴趣的:(树dp,树dp)