2019杭电多校第九场
熟悉的后半场挂机节奏，又苟进首页了，很快乐
1001. Rikka with Quicksort
upsolved
不是我做的,1e9调和级数分段打表
1002. Rikka with Cake
solved at 01:11
有一个矩形，给你很多射线（射线只有横平竖直的四个方向），问把矩形切成了多少块
队友说答案是交点数加一，作为一个合格的工具人，当然是把队友的想法实现啦
二维坐标离散化枚举纵坐标维护横坐标，常规套路，树状数组也可以做（我是线段树写习惯了根本没想起来还有树状数组）
#include <bits/stdc++.h>
using namespace std;
const int N = 1e5 + 10;
struct P {
int x, y;
char op[3];
}a[N];
long long ans;
int b[N], totx, toty;
int T, n, m, K;
vector<int> in[N], out[N], le[N], ri[N];
int sum[N << 2];
void pushup(int rt) {
sum[rt] = sum[rt << 1] + sum[rt << 1 | 1];
}
void build(int rt, int l, int r) {
if(l == r) {
sum[rt] = 0;
return;
}
int mid = l + r >> 1;
build(rt << 1, l, mid);
build(rt << 1 | 1, mid + 1, r);
pushup(rt);
}
void update(int rt, int l, int r, int pos, int val) {
if(l == r) {
sum[rt] += val;
return;
}
int mid = l + r >> 1;
if(pos <= mid) update(rt << 1, l, mid, pos, val);
else update(rt << 1 | 1, mid + 1, r, pos, val);
pushup(rt);
}
int query(int rt, int l, int r, int L, int R) {
if(L <= l && r <= R)
return sum[rt];
int mid = l + r >> 1, ans = 0;
if(L <= mid) ans += query(rt << 1, l, mid, L, R);
if(R > mid) ans += query(rt << 1 | 1, mid + 1, r, L, R);
return ans;
}
int main() {
scanf("%d", &T);
while(T--) {
ans = 0;
scanf("%d%d%d", &n, &m, &K);
for(int i = 1; i <= K; ++i) {
scanf("%d%d%s", &a[i].x, &a[i].y, a[i].op);
b[i] = a[i].x;
}
sort(b + 1, b + K + 1);
totx = unique(b + 1, b + K + 1) - b - 1;
for(int i = 1; i <= K; ++i)
a[i].x = lower_bound(b + 1, b + totx + 1, a[i].x) - b;
for(int i = 1; i <= K; ++i)
b[i] = a[i].y;
sort(b + 1, b + K + 1);
toty = unique(b + 1, b + K + 1) - b - 1;
for(int i = 1; i <= K; ++i)
a[i].y = lower_bound(b + 1, b + toty + 1, a[i].y) - b;
build(1, 1, totx);
for(int i = 1; i <= K; ++i)
in[i].clear(), out[i].clear(), le[i].clear(), ri[i].clear();
for(int i = 1; i <= K; ++i) {
if(a[i].op[0] == 'U')
in[a[i].y].push_back(a[i].x);
else if(a[i].op[0] == 'D') {
out[a[i].y].push_back(a[i].x);
update(1, 1, totx, a[i].x, 1);
}
else if(a[i].op[0] == 'L')
le[a[i].y].push_back(a[i].x);
else
ri[a[i].y].push_back(a[i].x);
}
for(int i = 1; i <= toty; ++i) {
for(auto f: in[i])
update(1, 1, totx, f, 1);
int l = 0, r = 1e9;
for(auto f: le[i])
l = max(l, f);
for(auto f: ri[i])
r = min(r, f);
if(l >= r)
ans += sum[1];
else {
if(l != 0)
ans += query(1, 1, totx, 1, l);
if(r != 1e9)
ans += query(1, 1, totx, r, totx);
}
//cout << "ans = " << ans << endl;
for(auto f: out[i])
update(1, 1, totx, f, -1);
}
printf("%lld\n", ans + 1);
}
return 0;
}1003. Rikka with Mista
upsolved
至多40个数，对每一个子集求其所有数的和的十进制表示里\(4\)的个数，对所有子集求和
先折半，然后按十进制位考虑，双指针查询（不用多次排序，只要在一次完整的基数排序的过程中计算就好了）
#include <bits/stdc++.h>
using namespace std;
using LL = long long;
struct num {
LL l, r;
};
LL ans, base;
vector<num> x, y, A[10], B[10];
int T, n, p, q, w[50];
void dfs(int now, int step, LL tmp, vector<num> &x) {
if(now == step) {
x.push_back({tmp, 0});
return;
}
dfs(now + 1, step, tmp, x);
dfs(now + 1, step, tmp + w[now], x);
}
LL get0(vector<num> &A, vector<num> &B, LL limit) {
LL res = 0;
int j = B.size() - 1;
for(int i = 0; i < A.size(); ++i) {
while(j >= 0 && A[i].r + B[j].r >= limit) j--;
res += j + 1;
}
return res;
}
LL get1(vector<num> &A, vector<num> &B, LL limit) {
LL res = 0;
int j = 0;
for(int i = (int)A.size() - 1; ~i; --i) {
while(j < B.size() && A[i].r + B[j].r < limit) j++;
res += (int)B.size() - j;
}
return res;
}
int main() {
scanf("%d", &T);
while(T--) {
ans = 0; base = 1;
x.clear(); y.clear();
scanf("%d", &n);
for(int i = 0; i < n; ++i)
scanf("%d", &w[i]);
p = n / 2;
dfs(0, p, 0, x);
dfs(p, n, 0, y);
for(int bit = 1; bit <= 9; ++bit) {
for(int i = 0; i <= 9; ++i)
A[i].clear(), B[i].clear();
for(auto f: x)
A[f.l % 10].push_back({f.l / 10, f.r});
for(auto f: y)
B[f.l % 10].push_back({f.l / 10, f.r});
for(int i = 0; i <= 9; ++i) {
int j1 = (14 - i) % 10, j2 = (13 - i) % 10;
ans += get0(A[i], B[j1], base) + get1(A[i], B[j2], base);
}
int nowx = 0, nowy = 0;
for(int i = 0; i <= 9; ++i) {
for(auto f: A[i])
x[nowx++] = (num){f.l, f.r + i * base};
for(auto f: B[i])
y[nowy++] = (num){f.l, f.r + i * base};
}
base *= 10;
}
printf("%lld\n", ans);
}
return 0;
}
1005. Rikka with Game
solved at 00:16
一个字符串，两个人轮流操作，每个人有两种操作，一是立即终止游戏，二是将一个位置的字符变成下一个字符\((a->b, b->c, ..., z->a)\)
第一个人想最小化字典序，第二个人想最大化字典序，求最后的字符串
想一想，发现是不考虑前缀\(y\), 如果首个字母是\(z\)则变成\(b\)，否则不变
1006. Rikka with Coin
solved at 00:51(+7)
有10,20,50,100四种面额的硬币，有\(n\)种商品，每种价格已知，求最少的硬币数使得硬币面额能恰好组成任意一种商品
设最贵的为\(w\)，面额100的要么是\(w/100\)个要么少一个，前三种暴力枚举
一开始max写成min了一直TLE...
#include <bits/stdc++.h>
using namespace std;
const int N = 110;
int T, n, w[N];
int ans, tmp;
int vis[22];
bool judge(int x, int y, int z, int d) {
memset(vis, 0, sizeof(vis));
vis[0] = 1;
for(int i = 1; i <= x; ++i) {
for(int j = 20; j >= 1; --j)
vis[j] |= vis[j - 1];
}
for(int i = 1; i <= y; ++i) {
for(int j = 20; j >= 2; --j)
vis[j] |= vis[j - 2];
}
for(int i = 1; i <= z; ++i) {
for(int j = 20; j >= 5; --j)
vis[j] |= vis[j - 5];
}
for(int i = 1; i <= n; ++i) {
bool flag = 0;
int t = w[i] % 10;
while(t <= 20) {
if(t + d * 10 >= w[i] && vis[t]) {flag = true; break;}
t += 10;
}
if(!flag) return false;
}
return true;
}
int main() {
scanf("%d", &T);
while(T--) {
bool flag = true;
scanf("%d", &n);
for(int i = 1; i <= n; ++i) {
scanf("%d", &w[i]);
if(w[i] % 10)
flag = false;
w[i] /= 10;
}
if(!flag) {
puts("-1");
continue;
}
sort(w + 1, w + n + 1);
tmp = w[n] / 10;
ans = 1e9;
for(int a = 0; a <= 10; ++a) {
for(int b = 0; b <= 5; ++b) {
for(int c = 0; c <= 2; ++c) {
for(int d = max(0, tmp - 1); d <= tmp; ++d) {
if(judge(a, b, c, d)) {
ans = min(ans, a + b + c + d);
}
}
}
}
}
printf("%d\n", ans);
}
return 0;
}1007. Rikka with Travel
solved at 02:13
给定一颗树，求满足存在一条点数为\(i\)的路径和一条点数为\(j\)的路径且两条路径不相交的点对\((i, j)\)的数量\((1<=n<=1e5)\)
我又是个工具人
树形dp
枚举一条边把树切成两半，然后两边分别求直径，假设为a, b, 那么\((1<=i<=a\&\&1<=j<=b)\)的点对都满足条件(\(i, j\)可以互换)
考虑两遍dfs, 第一遍处理出这个点的子树的直径，这个点往下延伸的最远的三个儿子以及长度，这个点的最大以及次大儿子直径
第二遍dfs处理出挖掉这个点的子树之后树的直径，可以用之前处理出的信息维护出来
#include <bits/stdc++.h>
using namespace std;
const int N = 1e5 + 10;
long long answer;
int v[N];
int T, n, x, y;
vector<int> G[N];
int mx[N][3], p[N][3], ans[N][2];
int mxson[N][2], ps[N][2];
void dfs(int rt, int fa) {
mx[rt][0] = 1; mx[rt][1] = -1e9; mx[rt][2] = -1e9;
p[rt][0] = rt; p[rt][1] = -1; mx[rt][2] = -1;
ans[rt][0] = 1;
mxson[rt][0] = mxson[rt][1] = -1e9;
ps[rt][0] = ps[rt][1] = -1;
for(int v: G[rt]) {
if(v == fa) continue;
dfs(v, rt);
if(mx[rt][2] < mx[v][0] + 1) {
mx[rt][2] = mx[v][0] + 1;
p[rt][2] = v;
}
if(mx[rt][2] > mx[rt][1]) {
swap(mx[rt][2], mx[rt][1]);
swap(p[rt][2], p[rt][1]);
}
if(mx[rt][1] > mx[rt][0]) {
swap(mx[rt][1], mx[rt][0]);
swap(p[rt][1], p[rt][0]);
}
if(ans[v][0] > mxson[rt][1]) {
mxson[rt][1] = ans[v][0];
ps[rt][1] = v;
}
if(mxson[rt][1] > mxson[rt][0]) {
swap(mxson[rt][1], mxson[rt][0]);
swap(ps[rt][1], ps[rt][0]);
}
ans[rt][0] = max(ans[rt][0], ans[v][0]);
}
ans[rt][0] = max(mx[rt][0] + mx[rt][1] - 1, ans[rt][0]);
}
void dfs2(int rt, int fa, int tmp, int len) {
ans[rt][1] = tmp;
for(int v: G[rt]) {
if(v == fa) continue;
int new_tmp = tmp;
int new_len = len + 1;
if(p[rt][0] != v) {
new_len = max(new_len, mx[rt][0]);
}
if(p[rt][1] != v) {
new_len = max(new_len, mx[rt][1]);
}
vector<int> vv;
vv.push_back(len + 1);
if(p[rt][0] != v)
vv.push_back(mx[rt][0]);
if(p[rt][1] != v)
vv.push_back(mx[rt][1]);
if(p[rt][2] != v)
vv.push_back(mx[rt][2]);
sort(vv.begin(), vv.end(), greater<int>());
new_tmp = max(new_tmp, vv[0] + vv[1] - 1);
if(ps[rt][0] != v)
new_tmp = max(new_tmp, mxson[rt][0]);
if(ps[rt][1] != v)
new_tmp = max(new_tmp, mxson[rt][1]);
dfs2(v, rt, new_tmp, new_len);
}
}
int main() {
scanf("%d", &T);
while(T--) {
answer = 0;
scanf("%d", &n);
for(int i = 1; i <= n; ++i) G[i].clear();
for(int i = 1; i < n; ++i) {
scanf("%d%d", &x, &y);
G[x].push_back(y);
G[y].push_back(x);
}
dfs(1, 0);
dfs2(1, 0, 0, 0);
memset(v, 0, sizeof(int) * (n + 3));
for(int i = 1; i <= n; ++i) {
int a = ans[i][0], b = ans[i][1];
v[a] = max(v[a], b);
v[b] = max(v[b], a);
}
for(int i = n; i; --i)
v[i] = max(v[i], v[i + 1]);
for(int i = 1; i <= n; ++i) {
answer += v[i];
}
printf("%lld\n", answer);
}
return 0;
}
1668
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
