Разбор
Tutorial is loading...
Решение (Vovuh)
#include <bits/stdc++.h>
using namespace std;
int main() {
#ifdef _DEBUG
freopen("input.txt", "r", stdin);
// freopen("output.txt", "w", stdout);
#endif
int n;
cin >> n;
vector<int> cnt(101);
for (int i = 0; i < n; ++i) {
int x;
cin >> x;
++cnt[x];
}
cout << *max_element(cnt.begin(), cnt.end()) << endl;
return 0;
}
1003B - Binary String Constructing
Разбор
Tutorial is loading...
Решение (Vovuh)
#include <bits/stdc++.h>
using namespace std;
int main() {
#ifdef _DEBUG
freopen("input.txt", "r", stdin);
// freopen("output.txt", "w", stdout);
#endif
int a, b, x;
cin >> a >> b >> x;
if (x % 2 == 0) {
if (a > b) {
for (int i = 0; i < x / 2; ++i)
cout << "01";
cout << string(b - x / 2, '1');
cout << string(a - x / 2, '0');
} else {
for (int i = 0; i < x / 2; ++i)
cout << "10";
cout << string(a - x / 2, '0');
cout << string(b - x / 2, '1');
}
} else if (a > b) {
for (int i = 0; i < x / 2; ++i)
cout << "01";
cout << string(a - x / 2, '0');
cout << string(b - x / 2, '1');
} else {
for (int i = 0; i < x / 2; ++i)
cout << "10";
cout << string(b - x / 2, '1');
cout << string(a - x / 2, '0');
}
cout << endl;
return 0;
}
Разбор
Tutorial is loading...
Решение (PikMike)
#include <bits/stdc++.h>
using namespace std;
#define x first
#define y second
#define mp make_pair
#define pb push_back
#define sqr(a) ((a) * (a))
#define sz(a) int(a.size())
#define all(a) a.begin(), a.end()
#define forn(i, n) for(int i = 0; i < int(n); i++)
#define fore(i, l, r) for(int i = int(l); i < int(r); i++)
typedef long long li;
typedef long double ld;
typedef pair<int, int> pt;
template <class A, class B> ostream& operator << (ostream& out, const pair<A, B> &a) {
return out << "(" << a.x << ", " << a.y << ")";
}
template <class A> ostream& operator << (ostream& out, const vector<A> &v) {
out << "[";
forn(i, sz(v)) {
if(i) out << ", ";
out << v[i];
}
return out << "]";
}
mt19937 rnd(time(NULL));
const int INF = int(2e9);
const li INF64 = li(1e18);
const int MOD = INF + 7;
const ld EPS = 1e-9;
const ld PI = acos(-1.0);
const int N = 10000 + 7;
int n, k;
int a[N];
bool read () {
if (scanf("%d%d", &n, &k) != 2)
return false;
forn(i, n) scanf("%d", &a[i]);
return true;
}
int pr[N];
void solve() {
pr[0] = 0;
forn(i, n) pr[i + 1] = pr[i] + a[i];
ld ans = 0;
forn(r, n) forn(l, r + 1){
if (r - l + 1 >= k){
ans = max(ans, (pr[r + 1] - pr[l]) / ld(r - l + 1));
}
}
printf("%.15f\n", double(ans));
}
int main() {
#ifdef _DEBUG
freopen("input.txt", "r", stdin);
// freopen("output.txt", "w", stdout);
int tt = clock();
#endif
cerr.precision(15);
cout.precision(15);
cerr << fixed;
cout << fixed;
while(read()) {
solve();
#ifdef _DEBUG
cerr << "TIME = " << clock() - tt << endl;
tt = clock();
#endif
}
}
Разбор
Tutorial is loading...
Решение (Vovuh)
#include <bits/stdc++.h>
using namespace std;
int main() {
#ifdef _DEBUG
freopen("input.txt", "r", stdin);
// freopen("output.txt", "w", stdout);
#endif
int n, q;
cin >> n >> q;
vector<int> cnt(31);
for (int i = 0; i < n; ++i) {
int x;
cin >> x;
++cnt[__builtin_ctz(x)];
}
while (q--) {
int x;
cin >> x;
int ans = 0;
for (int i = 30; i >= 0 && x > 0; --i) {
int need = min(x >> i, cnt[i]);
ans += need;
x -= (1 << i) * need;
}
if (x > 0)
ans = -1;
cout << ans << endl;
}
return 0;
}
Разбор
Tutorial is loading...
Решение (Vovuh)
#include <bits/stdc++.h>
using namespace std;
int main() {
#ifdef _DEBUG
freopen("input.txt", "r", stdin);
// freopen("output.txt", "w", stdout);
#endif
int n, d, k;
cin >> n >> d >> k;
if (d >= n) {
cout << "NO" << endl;
return 0;
}
vector<int> deg(n);
vector<pair<int, int>> ans;
set<pair<int, int>> q;
for (int i = 0; i < d; ++i) {
++deg[i];
++deg[i + 1];
if (deg[i] > k || deg[i + 1] > k) {
cout << "NO" << endl;
return 0;
}
ans.push_back(make_pair(i, i + 1));
}
for (int i = 1; i < d; ++i)
q.insert(make_pair(max(i, d - i), i));
for (int i = d + 1; i < n; ++i) {
while (!q.empty() && deg[q.begin()->second] == k)
q.erase(q.begin());
if (q.empty() || q.begin()->first == d) {
cout << "NO" << endl;
return 0;
}
++deg[i];
++deg[q.begin()->second];
ans.push_back(make_pair(i, q.begin()->second));
q.insert(make_pair(q.begin()->first + 1, i));
}
assert(int(ans.size()) == n - 1);
cout << "YES" << endl;
vector<int> p(n);
for (int i = 0; i < n; i++)
p[i] = i;
random_shuffle(p.begin(), p.end());
for (int i = 0; i < n - 1; ++i)
if (rand() % 2)
cout << p[ans[i].first] + 1 << " " << p[ans[i].second] + 1 << endl;
else
cout << p[ans[i].second] + 1 << " " << p[ans[i].first] + 1 << endl;
return 0;
}
Разбор
Tutorial is loading...
Решение (Vovuh)
#include <bits/stdc++.h>
using namespace std;
const int N = 303;
int n;
bool eq[N][N];
int dp[N][N];
string s[N];
int main() {
#ifdef _DEBUG
freopen("input.txt", "r", stdin);
// freopen("output.txt", "w", stdout);
#endif
cin >> n;
int allsum = n - 1;
for (int i = 0; i < n; ++i) {
cin >> s[i];
allsum += s[i].size();
}
for (int i = 0; i < n; ++i) {
eq[i][i] = true;
for (int j = 0; j < i; ++j) {
eq[i][j] = eq[j][i] = s[i] == s[j];
}
}
for (int i = n - 1; i >= 0; --i) {
for (int j = n - 1; j >= 0; --j) {
if (eq[i][j]) {
if (i + 1 < n && j + 1 < n)
dp[i][j] = dp[i + 1][j + 1] + 1;
else
dp[i][j] = 1;
}
}
}
int ans = allsum;
for (int i = 0; i < n; ++i) {
int sum = 0;
for (int j = 0; i + j < n; ++j) {
sum += s[i + j].size();
int cnt = 1;
for (int pos = i + j + 1; pos < n; ++pos) {
if (dp[i][pos] > j) {
++cnt;
pos += j;
}
}
int cur = allsum - sum * cnt + (j + 1) * cnt - j * cnt;
if (cnt > 1 && ans > cur) {
ans = cur;
}
}
}
cout << ans << endl;
return 0;
}
Отличный разбор. По мне сложность задачи B была выше чем C, D.
Мы тоже думали об этом, но в итоге решили, что B и C примерно одинаковой сложности. А вот с задачей D мы промахнулись, я не сразу увидел решение, которое потом описал в разборе, в итоге изначально авторское решение по ней было немного сложнее, чем сейчас, поэтому мы поставили ее на D.
Я долго думал как решить задачу B, после этого посмотрев задачу C изначально не поверил что оно так легко решается.
И какое было авторское решение?
Было все то же самое, только степени числа bj набирались каждая по отдельности и в массиве cnt хранился указатель, показывающий последнюю использовавшуюся степень, и число набиралось, можно, сказать, методом двух указателей.
Один из коротких решений для задачи B. 39962430
Прикольное:)
In E, won't the distance change upon inserting new edges and if yes, how to update the "maximal distance set" efficiently?
The previous explanation was wrong, i had understood it now.
If the distance for some vertex v will change after adding the new vertex, it means that we attach some leaf that increases the diameter of a tree. But it means that we attach a leaf to the vertex w with distw = d, but in this case we already doesn't have the answer.
It is possible to prove that there exists such a tree iff
n * (k - 2) <= -2 + k * (k - 1) ** (d // 2)
for even n andn * (k - 2) <= -2 + k * (k - 1) ** (d // 2) + (k - 2) * (k - 1) ** (d // 2)
for odd n. We simply check if conditions above are satisfied and proceed with making the tree or not.Edit: also, obviously it must be that
n > d
as said in the articleEdit:
k = 1
andk = 2
should be treated as special casesMy solution to F was to assign an integer to each distinct string (since there's a maximum of 300) while keeping track of lengths for each of these integers and then run a KMP search to find the number of occurrences of each valid subsequence. This solution runs in approximately the same time complexity as the editorial's.
My solution is same with you.39944563
Or maybe it can solved by hashing, but it may be hacked if I use hashing with overflow.39945601 39946612
There is an O(n) approach in problem C.
code
rough explanation
your pish and pop operations aren't constant time operations.
But each node would be push and pop 1 time.
I don't know why so much downvotes.
No idea why you're getting downvotes (that shouldn't matter tho, if you think your comment is constructive), your solution is very efficient.
It is a classic method to optimize DP.
I think I am doing nothing wrong to point it out in the case that editorial doesn't mention.
It seems that people are likely to downvote algorithm they don't know.
I think that is a very simple way to analysis the complexity, but it might be far too difficult for those rating is under 1600?
Can you please explain your solution?
what wrong in problem F makes a lotta participants output 689 in test 6 instead of 581?
Those solutions most likely did not consider more than two segments (only exactly two). The sample tests did not cover this.
can anyone explain me the method for Problem C becoz despite several efforts i m unable to figure it out
my approach is quite simple what I have done is to check for all segments the average temperature and we will consider which gives best answer and has
size >= k
my codeBro really thanks for ur help my approach is also quite the same but i got a TLE i m giving my code link plz help as I m beginner and really want to increase my skills http://codeforces.net/contest/1003/submission/39983625
I think "accumulate" as a third for loop .So the complexity is 5e3*5e3*5e3 , it will give you TLE. In worst case if n is 5e3 and k is 1 .
thanx bro ,,,, rest i think my algorithm is correct am I right??
You're right,but in this problem you don't have to use "accumulate".Think in a different way.
i fixed by using sum=sum+v[i];
Try it : ) . If it gives you a wrong answer , think about "what is the wrong in your code and what is the test you will fail in ", if you don't know , tell me , I will help you : ) .
thanx bro ...i passed all the test cases... thanx for ur help
In problem D my code passed all pretests but it timed out on 27th Case in system testing. No wonder how. Code complexity Q * 32 * 32. Can someone suggest something? Code
unordered_map operations works some more than O(1).
U can easily adjust to only Q * 32 operations with a greedy strategy. A complexity has no constants is it. your complexity is O(Q) if you say so. Map operations are O(log) so the number of operations is even worse in your solution. Just adapt to O(Q * log(MAX_VALUE) + N * log(MAX_VALUE)) which actually is O(N * log(VAX_VALUE)) because N and Q constrains are the same. Now to adapt you can just do this:
read editorial
Could someone explain why it didn't pass even the pretests? (problem D, WA test 4)
Code
for problem E,shortest code :http://codeforces.net/contest/1003/submission/39967116
why it times out on test 23?? (problem E) http://codeforces.net/contest/1003/submission/39989027 but if i assign the values of test 23 and upload the code here it will not time out!!: http://codeforces.net/contest/959/submission/39989086 why????????????? :(
Is it possible to solve problem D using Python?
Why is a random_shuffle performed at the end of the code for problem E?
Any O(nlogn) or O(n) method for C?
Не знаю насколько это ещё актуально, но, думаю, будет кому-то полезно.
1) E можно решать рекурсивно. Скажем
foo(x, dist)
— состояние рекурсии, когда мы уже добавили вершинуx
в наше дерево, при том, что текущий диаметр нашего дерева —dist
. Думаю, код будет достаточно компактным.2) В задаче F можно опустить ДП и решать с помощью Z-функции. Не так сильно полезно, но на какой-нибудь тренировочный контест можно дать именно как на Z-функцию.
what is wrong in my code for C ,it's giving the wrong answer on test#4 https://codeforces.net/contest/1003/submission/85153194
For problem E
My proof that when you attach a new node , d(x) of other nodes won't change.
Let d(x) be the distance of the furthest node from x, now when you attach a new node to node u, and say d(w) changes, then u is the furthest node from w (i.e. d(w) = dist(w,u)), so it means u is the end point of the diameter (d(u) = D, because that's the algorithm we use to find the diameter of a tree, finding the furthest node from a random node), so at that point we simple exit the process.