Tokitsukaze and a+b=n (easy)
Tokitsukaze and a+b=n (medium)
要使a+b=n,那么转换一下就是b=n-a,所以只需要计算[n-L,n-R]和[L',R']相交的部分即可
AC代码:
#include
using namespace std;
using LL = long long;
int main() {
ios::sync_with_stdio(false);
cin.tie(nullptr);
int T;
cin >> T;
while (T--) {
LL n;
cin >> n;
LL l1, l2, r1, r2;
cin >> l1 >> r1 >> l2 >> r2;
if (l1 > n || l2 > n) {
cout << "0\n";
continue;
}
if (r1 + r2 < n) {
cout << "0\n";
continue;
}
l1 = max(0LL, n - l1);
r1 = max(0LL, n - r1);
if (l1 > r1) {
swap(l1, r1);
}
if (l1 >= l2) {
if (l1 <= r2) {
cout << min(r2, r1) - l1 + 1 << '\n';
} else {
cout << "0\n";
}
} else {
if (r1 < l2) {
cout << "0\n";
} else {
cout << min(r1, r2) - l2 + 1 << '\n';
}
}
}
return 0;
}
Tokitsukaze and a+b=n (hard)
先运用差分数组c标记[L,R],再算差分数组的前缀和,即为每个点被区间覆盖了多少次,然后再算前缀和的前缀和,即为从1~i点一共被覆盖了多少次,计算有多少a+b=n,即计算[min(n-L,n-R),max(n-L,n-R)]=[L',R']区间覆盖次数,即为求presum[R']-presum[L'-1]再减去[L,R]和[L',R']重叠的部分
AC代码:
#include
using namespace std;
using LL = long long;
#define int long long
const int mod = 998244353;
int c[400010], presum[400010], presum1[400010];
signed main() {
ios::sync_with_stdio(false);
cin.tie(nullptr);
int n, m;
cin >> n >> m;
vector> a(m + 1);
for (int i = 1; i <= m; i++) {
cin >> a[i].first >> a[i].second;
c[a[i].first]++;
c[a[i].second + 1]--;
}
for (int i = 1; i <= 400005; i++) {
presum[i] = (presum[i - 1] + c[i]) % mod;
presum1[i] = (presum1[i - 1] + presum[i]) % mod;
}
LL ans = 0;
for (int i = 1; i <= m; i++) {
pair b;
b.first = n - a[i].first;
b.second = n - a[i].second;
if (b.first > b.second) {
swap(b.first, b.second);
}
if (b.second < 1) {
continue;
}
if (b.first < 1) {
b.first = 1;
}
if (a[i].second < b.first) {
} else if (a[i].first > b.second) {
} else if (a[i].first >= b.first && a[i].second <= b.second) {
ans = (ans - (a[i].second - a[i].first + 1) + mod) % mod;
} else if (b.first >= a[i].first && b.second <= a[i].second) {
ans = (ans - (b.second - b.first + 1) + mod) % mod;
} else if (b.first <= a[i].first && b.second <= a[i].second && a[i].first <= b.second) {
ans = (ans - (b.second - a[i].first + 1) + mod) % mod;
} else if (a[i].first <= b.first && a[i].second <= b.second && b.first <= a[i].second) {
ans = (ans - (a[i].second - b.first + 1) + mod) % mod;
}
ans = (ans + presum1[b.second] - presum1[b.first - 1] + mod) % mod;
}
cout << (ans + mod) % mod << '\n';
return 0;
}
Tokitsukaze and Energy Tree
根据题意贪心可知,能量越多的应该放在越深的节点上,这样到根的距离最长,计算次数最多
AC代码:
#include
using namespace std;
using LL = long long;
int main() {
ios::sync_with_stdio(false);
cin.tie(nullptr);
int n;
cin >> n;
vector> G(n + 1);
for (int i = 2; i <= n; i++) {
int u;
cin >> u;
G[i].push_back(u);
G[u].push_back(i);
}
vector a(n + 1), b(n + 1);
for (int i = 1; i <= n; i++) {
cin >> a[i];
}
sort(a.begin() + 1, a.end(), [&](LL x, LL y) {
return x > y;
});
vector> dep(n + 1);
function dfs = [&](int u, int fa) {
dep[u].first = dep[fa].first + 1;
dep[u].second = u;
for (auto v: G[u]) {
if (v != fa) {
dfs(v, u);
}
}
};
dfs(1, 0);
sort(dep.begin() + 1, dep.end(), [&](pair x, pair y) {
if (x.first != y.first) {
return x.first > y.first;
}
return x.second > y.second;
});
for (int i = 1; i <= n; i++) {
b[dep[i].second] = a[i];
}
vector sum(n + 1);
LL ans = 0;
function dfs1 = [&](int u, int fa) {
sum[u] = b[u];
for (auto v : G[u]) {
if (v != fa) {
dfs1(v, u);
sum[u] = sum[u] + sum[v];
}
}
ans = ans + sum[u];
};
dfs1(1, 0);
cout << ans << '\n';
return 0;
}
Tokitsukaze and Function
根据式子发现f(x)=n/x+x-1,可以令g(x)=n/x+x,根据均值不等式可知g(x)>=2sqrt(n),当且仅当n/x=x时成立,所以我们就找到了g(x)最小值的一个取值点,因为f(x)中n/x为下取整,所以根据数论分块可知,最小值会是一段,而不再是某一个点,因此需要判断sqrt(n),sqrt(n)+1是否在[l,r]中,因为g(x)为对勾函数,所以还需要判断两个端点l,r是否取值更小,然后就是固定左端点l,右端点为l,r,sqrt(n),sqrt(n+1)其中之一,二分查找最小的x使得f(x)最小
AC代码:
#include
using namespace std;
using LL = long long;
int main() {
ios::sync_with_stdio(false);
cin.tie(nullptr);
int T;
cin >> T;
while (T--) {
LL n, l, r;
cin >> n >> l >> r;
LL m = sqrt(n), m1 = sqrt(n) + 1;
LL pos = l, minn = 5e18;
if (n / m + m - 1 < minn && m >= l && r >= m) {
pos = m;
minn = n / m + m - 1;
}
if (n / m1 + m1 - 1 < minn && m1 >= l && r >= m1) {
pos = m1;
minn = n / m1 + m - 1;
}
if (n / l + l - 1 < minn) {
pos = l;
minn = n / l + l - 1;
}
if (n / r + r - 1 < minn) {
pos = r;
minn = n / r + r - 1;
}
LL L = l, R = pos;
while (L + 1 < R) {
LL mid = (L + R) >> 1;
if (n / mid + mid - 1 == minn) {
R = mid;
} else {
L = mid;
}
}
if (n / L + L - 1 == minn) {
cout << L << '\n';
} else {
cout << R << '\n';
}
}
return 0;
}
Tokitsukaze and Gold Coins (easy)
由题意可知需要保存k次操作后的地图,然后BFS从(1,1)跑一遍,标记能跑到的位置,再BFS从(n,3)跑一遍,标记能跑到的位置,如果(1,1)和(n,3)相互可达,那么两者都能跑到的即为能捡到的金币数
AC代码:
#include
using namespace std;
using LL = long long;
int dx[] = {1, 0};
int dy[] = {0, 1};
int dx1[] = {-1, 0};
int dy1[] = {0, -1};
int main() {
ios::sync_with_stdio(false);
cin.tie(nullptr);
int T;
cin >> T;
while (T--) {
int n, k;
cin >> n >> k;
vector> G(n + 1, vector (4));
for (int i = 1; i <= k; i++) {
int x, y;
cin >> x >> y;
G[x][y] ^= 1;
}
vector> vis(n + 1, vector (4)), vis1(n + 1, vector (4));
function check = [&](int x, int y) {
if (x < 1 || x > n || y < 1 || y > 3) {
return false;
}
return true;
};
queue> q;
q.push(make_pair(1, 1));
while (!q.empty()) {
pair now = q.front();
q.pop();
for (int i = 0; i < 2; i++) {
pair nxt;
nxt.first = now.first + dx[i];
nxt.second = now.second + dy[i];
if (check(nxt.first, nxt.second) && G[nxt.first][nxt.second] != 1 && !vis[nxt.first][nxt.second]) {
q.push(nxt);
vis[nxt.first][nxt.second] = 1;
}
}
}
if (!vis[n][3]) {
cout << "0\n";
continue;
}
q.push(make_pair(n, 3));
vis1[n][3] = true;
int ans = 0;
while (!q.empty()) {
pair now = q.front();
q.pop();
for (int i = 0; i < 2; i++) {
pair nxt;
nxt.first = now.first + dx1[i];
nxt.second = now.second + dy1[i];
if (check(nxt.first, nxt.second) && G[nxt.first][nxt.second] != 1 && !vis1[nxt.first][nxt.second]) {
q.push(nxt);
vis1[nxt.first][nxt.second] = 1;
}
}
}
if (!vis1[1][1]) {
cout << "0\n";
continue;
}
for (int i = 1; i <= n; i++) {
for (int j = 1; j <= 3; j++) {
if (vis1[i][j] && vis[i][j]) {
ans++;
}
}
}
cout << ans << '\n';
}
return 0;
}
Tokitsukaze and K-Sequence
根据题意,把序列分成k段,容易想到出现次数小于等于k的数均分到各个子序列中,每个数都能产生1点影响,那么出现次数大于k的数,贪心地考虑就是k-1个平均分到k-1个区间内,剩下多余的全部扔到最后一个区间内,产生的影响就是(k-1)
AC代码:
#include
using namespace std;
using LL = long long;
#define int long long
signed main() {
ios::sync_with_stdio(false);
cin.tie(nullptr);
int T;
cin >> T;
while (T--) {
int n;
cin >> n;
vector a(n + 1), cnt(n + 1);
map mp;
for (int i = 1; i <= n; i++) {
cin >> a[i];
mp[a[i]]++;
}
LL sum = 0, sum1 = 0, ans = 0;
for (auto it : mp) {
cnt[it.second]++;
}
for (int i = 1; i <= n; i++) {
sum += cnt[i];
}
for (int i = 1; i <= n; i++) {
sum1 += cnt[i];
ans += cnt[i] * i;
cout << ans + (sum - sum1) * (i - 1) << '\n';
}
}
return 0;
}
Tokitsukaze and Musynx
根据题意每个x[i]当且仅当有五种分数,而且也就只能发生4次分数的变化,所以一共是4*n,因此可以将x的坐标轴全部偏移到a的左边,这样所有的得分就是v[1],然后用multiset自动排序,pair记录,pair的key就是x[i]偏移到下个分数段的最左侧需要多少,val就是下个分数段和当前分数段的差值,因为排序是key从小到大排,那么肯定是从最先到达下各分数段的开始变化,所以不会出现x[i]到了下个分数段而有其他数跳过了多个分数段的情况,所以每次加上分数变化的差值,记录其中的最大值即可
AC代码:
#include
using namespace std;
using LL = long long;
int main() {
ios::sync_with_stdio(false);
cin.tie(nullptr);
int T;
cin >> T;
while (T--) {
int n;
cin >> n;
vector x(n + 1);
for (int i = 1; i <= n; i++) {
cin >> x[i];
x[i] -= 9e9;
}
LL a, b, c, d;
cin >> a >> b >> c >> d;
d++;
vector v(6);
for (int i = 1; i <= 5; i++) {
cin >> v[i];
}
LL ans = 1LL * n * v[1];
multiset> se;
for (int i = 1; i <= n; i++) {
se.insert(make_pair(a - x[i], v[2] - v[1]));
se.insert(make_pair(b - x[i], v[3] - v[2]));
se.insert(make_pair(c - x[i], v[4] - v[3]));
se.insert(make_pair(d - x[i], v[5] - v[4]));
}
LL res = ans;
for (auto it : se) {
ans += it.second;
res = max(res, ans);
}
cout << res << '\n';
}
return 0;
}
Tokitsukaze and Sum of MxAb
根据题目公式,如果两个数都大于0,那么结果就是a[i]+a[j],如果两个数都小于0,那么结果就是-a[i]-a[j],如果一个数a[i]大于0一个数a[j]小于0,结果就是a[i]-a[j],发现所有的数都会加到最终答案内,所以判断每个数被加了多少次即可,可知2n次
AC代码:
#include
using namespace std;
using LL = long long;
int main() {
ios::sync_with_stdio(false);
cin.tie(nullptr);
int T;
cin >> T;
while (T--) {
int n;
cin >> n;
vector a(n + 1);
LL ans = 0;
for (int i = 1; i <= n; i++) {
cin >> a[i];
if (a[i] < 0) {
a[i] = -a[i];
}
}
for (int i = 1; i <= n; i++) {
ans = ans + 1LL * (2 * n) * a[i];
}
cout << ans << '\n';
}
return 0;
}
Tokitsukaze and Synthesis and Traits
根据题意可知,给出一张无向无环图,每次询问就是问有多少条不同的边,可以知道的是除了菊花图极端情况下,一般每个点所能连出去的边都是不多的,大概平均是根号左右,菊花图不优就是不优在一个点连了非常多的边,如果暴力遍历这些点所连的边会耗时非常多,那么考虑把无向图变成有向图,为了预防菊花图这种情况,可以把度小的点指向度大的点,这样即使暴力遍历,最多也就根号n个点左右,而且也不妨碍答案的统计,时间复杂度
AC代码:
#include
using namespace std;
using LL = long long;
int main() {
ios::sync_with_stdio(false);
cin.tie(nullptr);
int n, m, q;
cin >> n >> m >> q;
vector> a(m);
vector grade(n + 1);
vector> G(n + 1);
for (int i = 0; i < m; i++) {
cin >> a[i].first >> a[i].second;
grade[a[i].first]++;
grade[a[i].second]++;
}
for (int i = 0; i < m; i++) {
if (grade[a[i].first] < grade[a[i].second]) {
G[a[i].first].push_back(a[i].second);
} else {
G[a[i].second].push_back(a[i].first);
}
}
while (q--) {
int k;
cin >> k;
vector b(k);
vector vis(n + 1);
for (int i = 0; i < k; i++) {
cin >> b[i];
vis[b[i]] = true;
}
int ans = 0;
for (int i = 0; i < k; i++) {
for (auto v : G[b[i]]) {
if (vis[v]) {
ans++;
}
}
}
cout << ans << '\n';
}
return 0;
}
Tokitsukaze and Three Integers
根据题意,需要计算的是(a[i]*a[j]+a[k])%p结果在0~p-1的次数,因为i,j,k都是1~n,n立方肯定凉透了,那么考虑n方log或者n方左右,那么可以提前预处理出a[i]*a[j]%p的结果在0~p-1的次数,又因为i,j,k各不相同,所以最终答案要减去(a[i]*a[j]+a[i])%p和(a[i]*a[j]+a[j])%p,这样i,j从1~n枚举常数太大,考虑减小常数,因为i,j都从1~n枚举,说明(i,j)都跑了两遍,所以每次减的时候要减2,计算+a[k]%p的时候同理
AC代码:
#include
using namespace std;
using LL = long long;
#define int long long
signed main() {
ios::sync_with_stdio(false);
cin.tie(nullptr);
int n, p;
cin >> n >> p;
vector a(n + 1), ans(p);
for (int i = 1; i <= n; i++) {
cin >> a[i];
a[i] %= p;
}
map mp1;
for (int i = 1; i <= n; i++) {
for (int j = i + 1; j <= n; j++) {
if (i != j) {
int x = a[i] * a[j] % p;
mp1[x]++;
x = (x + a[i]) % p;
ans[x] -= 2;
x = (x - a[i] + p) % p;
x = (x + a[j]) % p;
ans[x] -= 2;
}
}
}
for (auto it : mp1) {
for (int i = 1; i <= n; i++) {
ans[(it.first + a[i]) % p] += it.second * 2;
}
}
for (int i = 0; i < p; i++) {
cout << ans[i] << " \n"[i == p - 1];
}
return 0;
}