# TOTEM - Editorial

Contest

Author: Shahjalal Shohag
Tester: Ildar Gainullin
Editorialist: Rajarshi Basu

Easy-Medium

# PROBLEM:

You are given two integers N and M. You have to construct a matrix with N rows and M columns. Consider a multiset S which contains N + M integers: for each row and each column of this matrix, the MEX of the elements of this row/column belongs to S. Then, in the matrix you construct, the MEX of S must be maximum possible.

#### Constraints

• 1 \le T \le 400
• 1 \le N, M \le 100
• the sum of N \cdot M over all test cases does not exceed 10^5

# EXPLANATION:

WLOG assume that n\leq m. We need to make the mex of the rows and columns from 0 to K, so that the mex of S can be K+1. We want to maximise this value of K. Obviously, K can be at most N+M. Letâ€™s dive a bit deeper into the actual construction.

In the i^{th} column, letâ€™s try to make the mex =i. Then, we can use the numbers 0,1,2, \dots N, and exclude the number i. But we cannot make the mex greater than N using only columns. Letâ€™s see how such a construction till might look like:

Now lets try to fill in the rows so that we get some new Mex values. The following construction seems natural:

In these examples, n = 4 and m = 9. Now letâ€™s look at a similar construction but with n=4, m = 6 :

Here, whatever you place in the ? places, you wont get a mex greater than 6.

These examples lead to a simple formula : our best possible scenario is min(m,2*n) + 1. The construction method should also be clear from the diagrams. Some intuition might be that,

• by filling columns, we can get upto n values in |S|,
• the number of more values we can get by filling rows is min(m-n,n), that is, either we run out of rows (corresponding to n inside the min), or we run out of columns to make values higher than m (corresponding to m-n inside the min since we have already formed m by filling in columns).

# SOLUTION:

Setterâ€™s Code
#include<bits/stdc++.h>
using namespace std;

int MEX(set<int> &se) {
int ans = 0;
while (se.find(ans) != se.end()) ans++;
return ans;
}
int yo(vector<vector<int>> &a) {
set<int> se;
int n = a.size(), m = a[0].size();
for (int i = 0; i < n; i++) {
set<int> cur;
for (int j = 0; j < m; j++) {
cur.insert(a[i][j]);
}
se.insert(MEX(cur));
}
for (int j = 0; j < m; j++) {
set<int> cur;
for (int i = 0; i < n; i++) {
cur.insert(a[i][j]);
}
se.insert(MEX(cur));
}
return MEX(se);
}
int32_t main() {
ios_base::sync_with_stdio(0);
cin.tie(0);
int t; cin >> t;
assert(1 <= t && t <= 400);
int sum = 0;
while (t--) {
int n, m; cin >> n >> m;
assert(1 <= n && n <= 100);
assert(1 <= m && m <= 100);
sum += n * m;

bool f = 0;
if (n > m) {
swap(n, m);
f = 1;
}
int mx = 0;
while (1) {
if (mx <= n);
else if (mx - n <= n && mx <= m);
else break;
mx++;
}
if (n == 1 && m == 1) mx = 1;
vector<vector<int>> a(n, vector<int>(m, 0));
for (int i = 0; i < n; i++) {
for (int j = 0; j < m; j++) {
a[i][j] = (j - i + m) % m;
}
}
for (int i = 0, j = n + 1; i < n && j < m; i++, j++) a[i][(j + i) % m] = 0;
if (n == m) {
for (int j = 0; j < m; j++) a[0][j] = n + 1;
}
if (n == 1) {
for (int j = 0; j < m; j++) a[0][j] = !j;
}
if (f) {
vector<vector<int>> b(m, vector<int>(n, 0));
for (int i = 0; i < n; i++) {
for (int j = 0; j < m; j++) {
b[j][i] = a[i][j];
}
}
swap(n, m);
a = b;
}
for (int i = 0; i < n; i++) {
for (int j = 0; j < m; j++) {
cout << a[i][j] << ' ';
}
cout << '\n';
}
assert(mx == yo(a));
}
assert(1 <= sum && sum <= 100000);
return 0;
}

Testerâ€™s Code
#include <cmath>
#include <functional>
#include <fstream>
#include <iostream>
#include <vector>
#include <algorithm>
#include <string>
#include <set>
#include <map>
#include <list>
#include <time.h>
#include <math.h>
#include <random>
#include <deque>
#include <queue>
#include <cassert>
#include <unordered_map>
#include <unordered_set>
#include <iomanip>
#include <bitset>
#include <sstream>
#include <chrono>
#include <cstring>

using namespace std;

typedef long long ll;

#ifdef iq
mt19937 rnd(228);
#else
mt19937 rnd(chrono::high_resolution_clock::now().time_since_epoch().count());
#endif

int main() {
#ifdef iq
freopen("a.in", "r", stdin);
#endif
ios::sync_with_stdio(0);
cin.tie(0);
auto solve = [&] (int n, int m) {
vector <vector <int> > b(n, vector <int> (m, 1));
if (n == 1 && m == 1) return b;
bool sw = false;
if (n > m) swap(n, m), sw = true;
int ans = n + min(m - n, n) + 1;
vector <vector <int> > a(n, vector <int> (m));
for (int i = 0; i < n; i++) {
int x = (m - i) % m;
for (int j = 0; j < m; j++) {
a[i][j] = x;
x++;
x %= m;
}
}
for (int i = 0; i < n && n + 1 + i < m; i++) {
a[i][(n + 1 + 2 * i) % m] = 0;
}
if (n == m) {
for (int i = 0; i < n; i++) {
for (int j = 0; j < n; j++) {
a[i][j] = ((i + j) % n);
}
}
for (int i = 0; i < n; i++) a[0][i] = n + 1;
}
if (!sw) {
return a;
} else {
for (int i = 0; i < m; i++) {
for (int j = 0; j < n; j++) {
b[i][j] = a[j][i];
}
}
return b;
}
};
auto cost = [&] (vector <vector <int> > s) {
int n = (int) s.size(), m = (int) s[0].size();
set <int> arr;
for (int i = 0; i < n; i++) {
set <int> q;
for (int j = 0; j < m; j++) {
q.insert(s[i][j]);
}
int y = 0;
while (q.count(y)) y++;
arr.insert(y);
}
for (int j = 0; j < m; j++) {
set <int> q;
for (int i = 0; i < n; i++) {
q.insert(s[i][j]);
}
int y = 0;
while (q.count(y)) y++;
arr.insert(y);
}
int z = 0;
while (arr.count(z)) z++;
return z;
};
/*
while (true) {
int n = rnd() % 100 + 1;
int m = rnd() % 100 + 1;
auto x = solve(n, m);
if (cost(x) != min(min(n, m) * 2, max(n, m)) + 1 && (n != 1 || m != 1)) {
cout << n << ' ' << m << endl;
cout << cost(x) << endl;
return 0;
}
// assert(cost(x) == min(min(n, m) * 2, max(n, m)) + 1);
}
*/
int t;
cin >> t;
while (t--) {
int n, m;
cin >> n >> m;
auto x = solve(n, m);
for (int i = 0; i < n; i++) {
for (int j = 0; j < m; j++) {
cout << x[i][j] << ' ';
}
cout << '\n';
}
}
}

2 Likes

Common!! can anyone Please provide a mathematical proof for the formula max(MEX) = min(m,2*n)+1. How example is leading to that?? @rajarshi_basu

2 Likes

some more intuition given

Can you provide me a testcase please, for which my solution getting WA. I generated a checker code too to find the mex value of the matrix and it satisfy the best possible scenario â€śmin(m, 2*n) + 1â€ť for all possible pair of n, m. But still my solution got WA verdict.

your output for n=m=1 is wrong

2 Likes

thanksâ€¦ Got itâ€¦ Unfortunately I didnâ€™t observe this case and it caused me negative rating today

How to prove this is optimal construction?

Hope that it will help.
Let assume, n<=m. the mex value from the columns will be in range 0 to n as each column contains at most n elements. it will better to get all first n mex value {0, 1, â€¦, n} from columns. And we can get n more mex value {n+1, n+2, â€¦, m} from n row. thus we can get at most n+n distinct mex value from all row and column. But the maximum mex value for any row/column can be atmost m as a row can contain at most m elements. So, we can get at most min(n+n, m) distinct mex value from all row and column and the mex value of the matrix will be min(n+n, m)+1. by following the sequence of the editorial you can construct a matrix which will gives you the best answer which is min(n+n, m)+1.

3 Likes

why did i get tle for this??

#include <bits/stdc++.h>

using namespace std;

int main()
{
ios_base::sync_with_stdio(false);
cin.tie(NULL);cout.tie(NULL);

int t;
cin>>t;
while(t--)
{
int n,m;
cin>>n>>m;
int mat[n][m],i,j;

int cnt=1;

if(n>=m)
{
for(j=0;j<1;++j)
{
for(i=0;i<n;++i)
{
mat[i][j]=cnt;
cnt++;
}
}

cnt=0;

for(j=1;j<m;++j)
{
for(i=0;i<n;++i)
{
mat[i][j]=cnt;
cnt++;
}
}

for(i=0;i<n;++i)
{
for(j=0;j<m;++j)
{
cout<<mat[i][j]<<" ";

}

cout<<endl;
}
}

else
{
for(i=0;i<1;++i)
{
for(j=0;j<m;++j)
{
mat[i][j]=cnt;
cnt++;
}
}

cnt=0;

for(i=1;i<n;++i)
{
for(j=0;i<m;++j)
{
mat[i][j]=cnt;
cnt++;
}
}

for(i=0;i<n;++i)
{
for(j=0;j<m;++j)
{
cout<<mat[i][j]<<" ";

}

cout<<endl;
}
}
}

return 0;


}

Does anyone have a video tutorial on this? Iâ€™m not able to understand how to fill the rows and col

if n==m ans would be n,
if(n>m)
{
ans would be between (n,m),because if(mex==x) we need atleast x elements(0 to x-1).
so ans would be m for m rows;
and remaining rows in best case contributes min(m,n-m) elements(i.e,every column contributes one element).
we have m+min(m,n-m) elements.so ans is m+min(m,n-m)
}
vice versa for(n<m)

Can anybody help me to find the mistake in my code.

import java.util.;
import java.io.
;
class Mex
{
public static void main(String args[])
{
Scanner sc=new Scanner(System.in);
PrintWriter pw=new PrintWriter(System.out);
int T=sc.nextInt();
for(int i=0;i<T;i++)
{
int N=sc.nextInt();
int M=sc.nextInt();
int min=Math.min(N,M);
int arr[][]=new int[min][Math.max(N,M)];
int j;
for(j=0;j<min;j++)
{
for(int k=j;k>=0;kâ€“)
{
arr[k][j]=j-k;
}
}
if(N!=M)
j=min;
else
j=min-1;
for(int k=min-1;k>=0;kâ€“)
{
arr[k][j]=min-k;
}
if(j==min-1)
arr[0][Math.max(N,M)-1]=min-1;
for(int x=1;x<Math.min(N,M);x++)
{
for(int k=0;k<x;k++)
{
arr[x][k]=arr[x][j]+k+1;
}
}
int start=min+1,c=0;
for(int x=0;x<Math.min(Math.abs(N-M),min);x++)
{
start=min+1+x;
c=j+1;
for(int k=j+1;k<Math.max(N,M);k++)
{
if(start==c)
arr[x][k]=++c;
else
arr[x][k]=c;
c++;
}
}
if(N<=M)
{
for(int x=0;x<N;x++)
{
for(int y=0;y<M;y++)
{
System.out.print(arr[x][y]+" â€ś);
}
System.out.println();
}
}
else
{
for(int x=0;x<N;x++)
{
for(int y=0;y<M;y++)
{
System.out.print(arr[y][x]+â€ť ");
}
System.out.println();
}
}
}
pw.flush();
}
}

matrix can be filled by dividing each row into 3 parts and some special care when n=m
arr = new int[n][m];
// we divide every row into 3
// 0 to i , i to n and n to m

    for(int i = 0 ; i < n ; i++)
{

int last  = (m==n) ? n : n-i+1;

for(int j = 0 ; j < i ; j++)
{
arr[i][j] = last;
last++;
}

int c = 0;
int end  = (m==n) ? n-1 : n;
for(int j = i ; j <= end; j++)
{
arr[i][j] = c;
c++;
}

// value to be skipped is n+i+1
c = n+1;
for(int j = n+1 ; j <  m ; j++)
{
if( c == n+i+1 )
{
c++;
}
arr[i][j] = c;
c++;
}
}


if(n== m)
arr[n-1][m-1] = n*m;
// getting zero from last row

there was a video discussion yesterday where I discussed all the problems. Check it out on Codechefâ€™s channel on youtube

can someone help me in finding mistake in my solution