# Codeforces Round #618 (Div. 1+2)

## A. Non-zero

time limit per test1 second memory limit per test256 megabytes

Guy-Manuel and Thomas have an array a of n integers $$[a_1, a_2, \dots, a_n]$$. In one step they can add 1 to any element of the array. Formally, in one step they can choose any integer index $$i (1 \le i \le n)$$ and do $$a_i := a_i + 1$$.

If either the sum or the product of all elements in the array is equal to zero, Guy-Manuel and Thomas do not mind to do this operation one more time.

What is the minimum number of steps they need to do to make both the sum and the product of all elements in the array different from zero? Formally, find the minimum number of steps to make $$a_1 + a_2 + \dots + a_n \ne 0 and a_1 \cdot a_2 \cdot \dots \cdot a_n \ne 0$$.Input

Each test contains multiple test cases.

The first line contains the number of test cases $$t (1 \le t \le 10^3)$$. The description of the test cases follows.

The first line of each test case contains an integer $$n (1 \le n \le 100)$$— the size of the array.

The second line of each test case contains n integers $$a_1, a_2, \dots, a_n (-100 \le a_i \le 100)$$ — elements of the array .Output

For each test case, output the minimum number of steps required to make both sum and product of all elements in the array different from zero.ExampleinputCopy

4
3
2 -1 -1
4
-1 0 0 1
2
-1 2
3
0 -2 1


outputCopy

1
2
0
2


Note

In the first test case, the sum is 0. If we add 1 to the first element, the array will be [3,-1,-1], the sum will be equal to 1 and the product will be equal to 3.

In the second test case, both product and sum are 0. If we add 1 to the second and the third element, the array will be [-1,1,1,1], the sum will be equal to 2 and the product will be equal to -1. It can be shown that fewer steps can't be enough.

In the third test case, both sum and product are non-zero, we don't need to do anything.

In the fourth test case, after adding 1 twice to the first element the array will be [2,-2,1], the sum will be 1 and the product will be -4.

#### 签到

#include<bits/stdc++.h>
using namespace std;
#define rep(i,a,b) for(int i=(a);i<=(b);++i)
#define dep(i,a,b) for(int i=(a);i>=(b);--i)
#define pb push_back
typedef long long ll;
const int maxn=(int)1e6+100;
const int mod=(int)1e9+7;
int n,a[maxn];
void solve(){
int ans=0;
scanf("%d",&n);
rep(i,1,n){
scanf("%d",&a[i]);
if(a[i]==0) a[i]++,ans++;
}
int sum=0;
rep(i,1,n) sum+=a[i];
if(sum==0) ans++;
printf("%d\n",ans);
}
int main(){
int T;cin>>T;
while(T--) solve();
}


## B. Assigning to Classes

time limit per test2 seconds memory limit per test256 megabytes

Reminder: the median of the array $$[a_1, a_2, \dots, a_{2k+1}]$$ of odd number of elements is defined as follows: let$$[b_1, b_2, \dots, b_{2k+1}]$$be the elements of the array in the sorted order. Then median of this array is equal to $$b_{k+1}$$.

There are 2n students, the i-th student has skill level $$a_i$$. It's not guaranteed that all skill levels are distinct.

Let's define skill level of a class as the median of skill levels of students of the class.

As a principal of the school, you would like to assign each student to one of the 2 classes such that each class has odd number of students (not divisible by 2). The number of students in the classes may be equal or different, by your choice. Every student has to be assigned to exactly one class. Among such partitions, you want to choose one in which the absolute difference between skill levels of the classes is minimized.

What is the minimum possible absolute difference you can achieve?Input

Each test contains multiple test cases. The first line contains the number of test cases $$t (1 \le t \le 10^4)$$. The description of the test cases follows.

The first line of each test case contains a single integer $$n (1 \le n \le 10^5)$$ — the number of students halved.

The second line of each test case contains 2n integers $$a_1, a_2, \dots, a_{2 n} (1 \le a_i \le 10^9)$$ — skill levels of students.

It is guaranteed that the sum of n over all test cases does not exceed $$10^5$$.Output

For each test case, output a single integer, the minimum possible absolute difference between skill levels of two classes of odd sizes.ExampleinputCopy

3
1
1 1
3
6 5 4 1 2 3
5
13 4 20 13 2 5 8 3 17 16


outputCopy

0
1
5


Note

In the first test, there is only one way to partition students — one in each class. The absolute difference of the skill levels will be |1 - 1| = 0.

In the second test, one of the possible partitions is to make the first class of students with skill levels [6, 4, 2], so that the skill level of the first class will be 4, and second with [5, 1, 3], so that the skill level of the second class will be 3. Absolute difference will be |4 - 3| = 1.

Note that you can't assign like [2, 3], [6, 5, 4, 1] or [], [6, 5, 4, 1, 2, 3] because classes have even number of students.

, [1, 3, 4] is also not possible because students with skills 5 and 6 aren't assigned to a class.

In the third test you can assign the students in the following way: [3, 4, 13, 13, 20], [2, 5, 8, 16, 17] or [3, 8, 17], [2, 4, 5, 13, 13, 16, 20]. Both divisions give minimal possible absolute difference.

#### 排序练习题

#include<bits/stdc++.h>
using namespace std;
#define rep(i,a,b) for(int i=(a);i<=(b);++i)
#define dep(i,a,b) for(int i=(a);i>=(b);--i)
#define pb push_back
typedef long long ll;
const int maxn=(int)1e6+100;
const int mod=(int)1e9+7;
int n,a[maxn];
void solve(){
scanf("%d",&n);
rep(i,1,2*n) scanf("%d",&a[i]);
sort(a+1,a+1+2*n);
printf("%d\n",a[n+1]-a[n]);
}
int main(){
int T;cin>>T;
while(T--) solve();
}


## C. Anu Has a Function

time limit per test1 second memory limit per test256 megabytes

Anu has created her own function f: f(x, y) = (x | y) - y where | denotes the bitwise OR operation. For example, f(11, 6) = (11|6) - 6 = 15 - 6 = 9. It can be proved that for any nonnegative numbers x and y value of f(x, y) is also nonnegative.

She would like to research more about this function and has created multiple problems for herself. But she isn't able to solve all of them and needs your help. Here is one of these problems.

A value of an array $$[a_1, a_2, \dots, a_n]$$ is defined as $$f(f(\dots f(f(a_1, a_2), a_3), \dots a_{n-1}), a_n)$$ (see notes). You are given an array with not necessarily distinct elements. How should you reorder its elements so that the value of the array is maximal possible?Input

The first line contains a single integer $$n (1 \le n \le 10^5)$$.

The second line contains n integers $$a_1, a_2, \ldots, a_n (0 \le a_i \le 10^9)$$. Elements of the array are not guaranteed to be different.Output

Output n integers, the reordering of the array with maximum value. If there are multiple answers, print any.ExamplesinputCopy

4
4 0 11 6


outputCopy

11 6 4 0

inputCopy

1
13


outputCopy

13

Note

In the first testcase, value of the array [11, 6, 4, 0] is f(f(f(11, 6), 4), 0) = f(f(9, 4), 0) = f(9, 0) = 9.

[11, 4, 0, 6] is also a valid answer.

#### 显而易见，f(x,y)的作用就是 在x的二进制中，把所有y中为1的位置置为0；那么显然除了第一个数，后面的n-1个数的顺序是没影响的，那我们就需要去决定第一个数；有了上面的性质，显然我们只要找一个只有一个数为1的最高的二进制位，这个位对应的数就是第一个数，后面随便输就好了

#include<bits/stdc++.h>
using namespace std;
#define rep(i,a,b) for(int i=(a);i<=(b);++i)
#define dep(i,a,b) for(int i=(a);i>=(b);--i)
#define pb push_back
typedef long long ll;
const int maxn=(int)1e6+100;
const int mod=(int)1e9+7;
int n,a[maxn],cnt,now;
int main(){
scanf("%d",&n);
rep(i,1,n){
scanf("%d",&a[i]);int x=a[i],pos=0;
while(x){
cnt[pos++]+=(x&1);
if(x&1) now[pos-1]=i;
x>>=1;
}
}
ll ans=0,p=-1;
rep(i,0,62) if(cnt[i]==1) ans=(1ll<<i),p=now[i];
if(p!=-1) printf("%d ",a[p]);
rep(i,1,n) if(i!=p) printf("%d ",a[i]);
}


## D. Aerodynamic

time limit per test1 second memory limit per test256 megabytes

Guy-Manuel and Thomas are going to build a polygon spaceship.

You're given a strictly convex (i. e. no three points are collinear) polygon P which is defined by coordinates of its vertices. Define P(x,y) as a polygon obtained by translating P by vector $$\overrightarrow {(x,y)}$$. The picture below depicts an example of the translation:

Define T as a set of points which is the union of all P(x,y) such that the origin (0,0) lies in P(x,y) (both strictly inside and on the boundary). There is also an equivalent definition: a point (x,y) lies in T only if there are two points A,B in P such that $$\overrightarrow {AB} = \overrightarrow {(x,y)}$$. One can prove T is a polygon too. For example, if P is a regular triangle then T is a regular hexagon. At the picture below P is drawn in black and some P(x,y) which contain the origin are drawn in colored:

The spaceship has the best aerodynamic performance if P and T are similar. Your task is to check whether the polygons P and T are similar.Input

The first line of input will contain a single integer $$n (3 \le n \le 10^5)$$ — the number of points.

The i-th of the next n lines contains two integers$$x_i, y_i (|x_i|, |y_i| \le 10^9)$$, denoting the coordinates of the i-th vertex.

It is guaranteed that these points are listed in counterclockwise order and these points form a strictly convex polygon.Output

Output "YES" in a separate line, if P and T are similar. Otherwise, output "NO" in a separate line. You can print each letter in any case (upper or lower).ExamplesinputCopy

4
1 0
4 1
3 4
0 3


outputCopy

YES

inputCopy

3
100 86
50 0
150 0


outputCopy

nO

inputCopy

8
0 0
1 0
2 1
3 3
4 6
3 6
2 5
1 3


outputCopy

YES

Note

The following image shows the first sample: both P and T are squares. The second sample was shown in the statements.

#### 这道题难点在题意；依次选择多边形上的点 平移这个多边形 使得选中的点在原点上，那么这时候剩下n-1个点，重复n次，那么一共有n*(n-1)个点，问你这些点构成的多边形和原来的是不是相似那其实就是判前n/2条边和后n/2条边是否平行且相等

#include<bits/stdc++.h>
using namespace std;
#define rep(i,a,b) for(int i=(a);i<=(b);++i)
#define dep(i,a,b) for(int i=(a);i>=(b);--i)
#define pb push_back
typedef long long ll;
const int maxn=(int)1e6+100;
const int mod=(int)1e9+7;
int n;
struct node{
ll x,y;
}a[maxn],p[maxn];
int main(){
scanf("%d",&n);
rep(i,1,n) scanf("%lld%lld",&a[i].x,&a[i].y);
a[n+1].x=a.x;a[n+1].y=a.y;
if(n%2) return puts("NO"),0;
rep(i,1,n){
p[i].x=a[i+1].x-a[i].x;
p[i].y=a[i+1].y-a[i].y;
}
rep(i,1,n/2) if(p[i].x!=-p[i+n/2].x||p[i].y!=-p[i+n/2].y) return puts("NO"),0;
puts("YES");
}


## E. Water Balance

time limit per test3 seconds memory limit per test256 megabytes

There are n water tanks in a row, i-th of them contains $$a_i$$ liters of water. The tanks are numbered from 1 to n from left to right.

You can perform the following operation: choose some subsegment [l, r] $$(1\le l \le r \le n)$$, and redistribute water in tanks $$l, l+1, \dots, r$$ evenly. In other words, replace each of $$a_l, a_{l+1}, \dots, a_r by \frac{a_l + a_{l+1} + \dots + a_r}{r-l+1}$$. For example, if for volumes [1, 3, 6, 7] you choose l = 2, r = 3, new volumes of water will be [1, 4.5, 4.5, 7]. You can perform this operation any number of times.

What is the lexicographically smallest sequence of volumes of water that you can achieve?

As a reminder:

A sequence a is lexicographically smaller than a sequence b of the same length if and only if the following holds: in the first (leftmost) position where a and b differ, the sequence a has a smaller element than the corresponding element in b.Input

The first line contains an integer $$n (1 \le n \le 10^6)$$ — the number of water tanks.

The second line contains n integers $$a_1, a_2, \dots, a_n (1 \le a_i \le 10^6)$$ — initial volumes of water in the water tanks, in liters.

Because of large input, reading input as doubles is not recommended.Output

Print the lexicographically smallest sequence you can get. In the i-th line print the final volume of water in the i-th tank.

Your answer is considered correct if the absolute or relative error of each $$a_i$$ does not exceed $$10^{-9}$$.

Formally, let your answer be $$a_1, a_2, \dots, a_n$$, and the jury's answer be $$b_1, b_2, \dots, b_n$$. Your answer is accepted if and only if $$\frac{|a_i - b_i|}{\max{(1, |b_i|)}} \le 10^{-9}$$ for each i.ExamplesinputCopy

4
7 5 5 7


outputCopy

5.666666667
5.666666667
5.666666667
7.000000000


inputCopy

5
7 8 8 10 12


outputCopy

7.000000000
8.000000000
8.000000000
10.000000000
12.000000000


inputCopy

10
3 9 5 5 1 7 5 3 8 7


outputCopy

3.000000000
5.000000000
5.000000000
5.000000000
5.000000000
5.000000000
5.000000000
5.000000000
7.500000000
7.500000000


Note

In the first sample, you can get the sequence by applying the operation for subsegment [1, 3].

In the second sample, you can't get any lexicographically smaller sequence.

#### 思维题，要求字典序最小，我们假设在末尾加上一个a[i]，并且a[i]<=a[i-1]，那么显然将这个a[i]与前面合并会使得答案更好，直到合并到此时的a[i]最大，那显然我们在单调栈中维护两个值，一个是这一段的平均值，另一个是这一段的长度

#include<bits/stdc++.h>
using namespace std;
#define rep(i,a,b) for(int i=(a);i<=(b);++i)
#define dep(i,a,b) for(int i=(a);i>=(b);--i)
#define pb push_back
typedef long long ll;
const int maxn=(int)1e6+100;
const int mod=(int)1e9+7;
int n,t[maxn],pos;
double a[maxn];
int main(){
scanf("%d",&n);
rep(i,1,n){
int x;scanf("%d",&x);
a[++pos]=x;t[pos]=1;
while(pos>1&&a[pos]<=a[pos-1]){
pos--;
a[pos]=(a[pos+1]*t[pos+1]+a[pos]*t[pos])/(t[pos]+t[pos+1]);
t[pos]+=t[pos+1];
}
}
rep(i,1,pos) rep(j,1,t[i]) printf("%.10lf\n",a[i]);
}


0 评论