zoukankan      html  css  js  c++  java
  • P2486 [SDOI2011]染色

    P2486

    很经典的题~

    思路: 线段树染色+"熟练"剖分(某些出题人总是喜欢把序列上的题加个树链剖分搞到树上去)

    先想一想序列上怎么做吧

    线段树是个好东西

    每个节点维护三个信息: ls: 左端点的颜色 rs: 右端点的颜色 cnt: [l, r] 中共有几个颜色段

    合并?

    fa.cnt = son1.cnt + son2.cnt - [son1.rs == son2.ls]

    fa.ls = son1.ls , fa.rs = son2.rs

    爹的左端点颜色就是左儿子的左端点颜色, 右端点亦然

    如果左儿子与右儿子相接的颜色相同, 那么等于左儿子块数加右儿子块数-1(中间两个块会合成一个)

    否则直接加就行啦

    修改时要打标记 记录有没有被覆盖

    回到树上问题时要特别注意的是询问

    因为询问时有swap的操作, 将k记录x,y的顺序, 即相当于(x, y) 还是(y, x)

    如果是(y, x), 最后还要反回来才能进行合并

    在跳重链时, 总是将链接在它的左边, 最后将a左右儿子反一下再与b合并即可

    #include<iostream>
    #include<cstdio>
    #include<cstdio>
    #define ll long long
    using namespace std;
    const int N = 105000*4;
    int fa[N], id[N], siz[N];
    int num, dep[N], son[N];
    int w[N], wt[N], Top[N];
    int h[N], ne[N], to[N];
    int tot;
    inline void add(int x,int y) {
    	ne[++tot] = h[x], h[x] = tot;
    	to[tot] = y;
    }
    void dfs1(int x,int f) {
    	fa[x] = f;
    	siz[x] = 1, dep[x] = dep[f] + 1;
    	for (int i = h[x]; i ;i = ne[i]) {
    		int y = to[i];
    		if (y == f) continue;
    		dfs1(y, x);
    		siz[x] += siz[y];
    		if (siz[y] > siz[son[x]]) son[x] = y;
    	}
    }
    void dfs2(int x,int topf) {
    	id[x] = ++num; wt[num] = w[x];
    	Top[x] = topf;
    	if (!son[x]) return;
    	dfs2(son[x], topf);
    	for (int i = h[x]; i; i = ne[i]) {
    		int y = to[i];
    		if (y == fa[x] || y == son[x]) continue;
    		dfs2(y, y);
    	}
    }
    int n, m;
    int L[N], R[N], cnt[N], ls[N], rs[N];
    int tag[N];
    #define p1 p << 1
    #define p2 p << 1 | 1
    
    struct node{
    	int cnt, ls, rs;
    };
    void update(node &fa,node i,node j) {
    	fa.cnt = i.cnt + j.cnt - (i.rs == j.ls);
    	fa.ls = i.ls, fa.rs = j.rs;
    }
    
    void build(int l,int r,int p) {
    	L[p] = l, R[p] = r;
    	if (l == r) {
    		cnt[p] = 1, ls[p] = rs[p] = wt[l];
    		return;
    	}
    	int mid = (l + r) >> 1;
    	build(l, mid, p1);
    	build(mid+1, r, p2);
    	cnt[p] = cnt[p1] + cnt[p2] - (rs[p1] == ls[p2]);
    	ls[p] = ls[p1], rs[p] = rs[p2];
    }
    
    void spread(int p) {
    	if (tag[p]) {
    		cnt[p1] = cnt[p2] = 1;
    		tag[p1] = tag[p2] = tag[p];
    		ls[p1] = ls[p2] = rs[p1] = rs[p2] = tag[p];
    		tag[p] = 0;
    	}
    }
    
    void change(int l,int r,int p,int c) {
    	if (L[p] >= l && R[p] <= r) {
    		cnt[p] = 1, tag[p] = c;
    		ls[p] = rs[p] = c;
    		return;
    	}
    	spread(p);
    	if (R[p1] >= l) change(l, r, p1, c);
    	if (L[p2] <= r) change(l, r, p2, c);
    	cnt[p] = cnt[p1] + cnt[p2] - (rs[p1] == ls[p2]);
    	ls[p] = ls[p1], rs[p] = rs[p2];
    }
    
    node ask(int l,int r,int p) {
    	if (L[p] >= l && R[p] <= r) return (node){cnt[p], ls[p], rs[p]};
    	spread(p);
    	node i;
    	if (R[p1] < l) return ask(l, r, p2);
    	if (L[p2] > r) return ask(l, r, p1);
    	update(i, ask(l, r, p1), ask(l, r, p2));
    	return i;
    }
    	
    void change_e(int x,int y,int c) {
    	while (Top[x] != Top[y]) {
    		if (dep[Top[x]] < dep[Top[y]]) swap(x, y);
    		change(id[Top[x]], id[x], 1, c);
    		x = fa[Top[x]];
    	}
    	if (dep[x] < dep[y]) swap(x, y);
    	change(id[y], id[x], 1, c);
    }
    
    int sum(int x,int y) {
    	node ans, a, b;
    	ans = a = b = (node){0,0,0};
    	int k = 1;
    	while (Top[x] != Top[y]) {
    		if (dep[Top[x]] < dep[Top[y]]) swap(x, y), swap(a, b), k ^= 1;
    		if (a.cnt == 0) a = ask(id[Top[x]], id[x], 1);
    		else update(a, ask(id[Top[x]], id[x], 1), a);
    		x = fa[Top[x]];
    	}
    	if (dep[x] < dep[y]) swap(x, y), swap(a, b);
    	if (a.cnt == 0) a = ask(id[y], id[x], 1);
    	else update(a, ask(id[y], id[x], 1), a);
    	if (b.cnt == 0) return a.cnt;
    	if (a.cnt == 0) return b.cnt;
    	if (!k) swap(a, b); // 将a, b恢复正常顺序
    	swap(a.ls, a.rs); //将a左右儿子换位
    	update(ans, a, b);
    	return ans.cnt;
    }
    	
    		
    char s[5];
    ll a, b, c;
    
    template <typename T> 
    void read(T &x) {
    	x = 0; int f = 1;
    	char c = getchar();
    	for (;!isdigit(c); c = getchar()) if (c == '-') f = -1;
    	for (;isdigit(c); c = getchar()) x = (x << 3) + (x << 1) + c - '0';
    	x *= f;
    }
    int main() {
    	read(n), read(m);
    	for (int i = 1;i <= n; i++) read(w[i]);
    	for (int i = 1;i < n; i++) {
    		read(a), read(b);
    		add(a, b); add(b, a);
    	}
    	dfs1(1, 0); 
    	dfs2(1, 1);
    	build(1, n, 1);
    	while (m--) {
    		scanf ("%s", s + 1);
    		if (s[1] == 'C') {
    			read(a), read(b), read(c);
    			change_e(a, b, c);
    		}
    		else {
    			read(a), read(b);
    			printf ("%d
    ", sum(a, b));
    		}
    	}
    	return 0;
    }
    
  • 相关阅读:
    机器学习项目实战----信用卡欺诈检测(二)
    机器学习项目实战----信用卡欺诈检测(一)
    机器学习之使用Python完成逻辑回归
    逻辑回归算法
    机器学习之梯度下降
    线性回归算法原理推导
    Spark之RDD容错原理及四大核心要点
    Spark之RDD依赖关系及DAG逻辑视图
    Spark之RDD弹性特性
    Spark之RDD的定义及五大特性
  • 原文地址:https://www.cnblogs.com/Hs-black/p/11645207.html
Copyright © 2011-2022 走看看