Maximum Sum possible by selecting X elements from a Matrix based on given conditions

Given a matrix G[][] of dimensions N × M, consists of positive integers, the task is to select X elements from the matrix having maximum sum considering the condition that G[i][j] can only be selected from the matrix unless all the elements G[i][k] are selected, where 0 ≤ k < j i.e., the jth element in the current ith row can be selected all its the preceding elements of the current ith row has already been selected.

Examples:

Input: N = 4, M = 4, X = 6, G[][] = {{3, 2, 6, 1}, {1, 9, 2, 4}, {4, 1, 3, 9}, {3, 8, 2, 1}} 
Output: 28 
Explanation: 
Selecting the first element from the 1st row = 3 
Selecting the the first two elements from the 2nd row = 1 + 9 = 10 
Selecting the first element from the 3rd row = 4 
Selecting the first two elements from the 4th row = 3 + 8 = 11 
Hence, the selected elements are {G[0][0], G[1][0], G[1][1], G[2][0], G[3][0], G[3][1]} 
Hence, the sum of the selected elements is = 3 + 10 + 4 + 11 = 28

Input: N = 2, M = 4, X = 4, G[][] = {{10, 10, 100, 30}, {80, 50, 10, 50}} 
Output: 200

Naive Approach: The simplest approach to solve this problem is to calculate the sum for all possible M selections and find the maximum sum among them.



Time Complexity: O(NM)
Auxiliary Space: O(1)

Efficient Approach: The above approach can be optimized using Dynamic Programming. The considerable states are: 

  1. Number of rows selected: i.
  2. Number of elements selected: j.

Initialize a matrix dp[][] such that dp[i][j] stores the maximum possible sum that can be obtained by selecting j elements from the first i rows.

The transition for the dp[][] is as follows:

dp[i][j] = maxx=(0, min(j, m))(dp[i – 1][j – x] + prefsum[i][x]) 
where prefsum[i][x] is the sum of the first x elements in the ith row of the matrix.

Below is the implementation of above approach:

C++14

filter_none

edit
close

play_arrow

link
brightness_4
code

// C++14 program to implement 
// the above approach 
#include <bits/stdc++.h>
using namespace std;
 
int n, m, X;
 
// Function to calculate the maximum
// possible sum by selecting X elements
// from the Matrix
int maxSum(vector<vector<int>> grid)
{
     
    // Generate prefix sum of the matrix
    vector<vector<int>> prefsum(n, vector<int>(m));
 
    for(int i = 0; i < n; i++)
    {
        for(int x = 0; x < m; x++)
        {
            if (x == 0)
                prefsum[i][x] = grid[i][x];
            else
                prefsum[i][x] = prefsum[i][x - 1] +
                                   grid[i][x];
        }
    }
 
    vector<vector<int>> dp(n, vector<int>(X + 1, INT_MIN));
 
    // Maximum possible sum by selecting
    // 0 elements from the first i rows
    for(int i = 0; i < n; i++)
           dp[i][0] = 0;
 
    // If a single row is present
    for(int i = 1; i <= min(m, X); ++i)
    {
        dp[0][i] = dp[0][i - 1] +
                 grid[0][i - 1];
    }
 
    for(int i = 1; i < n; ++i)
    {
        for(int j = 1; j <= X; ++j)
        {
             
            // If elements from the
            // current row is not selected
            dp[i][j] = dp[i - 1][j];
 
            // Iterate over all possible
            // selections from current row
            for(int x = 1; x <= min(j, m); x++)
            {
                dp[i][j] = max(dp[i][j],
                               dp[i - 1][j - x] +
                              prefsum[i][x - 1]);
            }
        }
    }
     
    // Return maximum possible sum
    return dp[n - 1][X];
}
 
// Driver code
int main()
{
    n = 4;
    m = 4;
    X = 6;
     
    vector<vector<int>> grid = { { 3, 2, 6, 1 },
                                 { 1, 9, 2, 4 },
                                 { 4, 1, 3, 9 },
                                 { 3, 8, 2, 1 } };
     
    int ans = maxSum(grid);
     
    cout << (ans);
    return 0;
}
 
// This code is contributed by mohit kumar 29

chevron_right


Java

filter_none

edit
close

play_arrow

link
brightness_4
code

// Java program to implement
// the above approach
import java.util.*;
import java.io.*;
 
class GFG {
 
    static int n, m, X;
 
    // Function to calculate the maximum
    // possible sum by selecting X elements
    // from the Matrix
    public static int maxSum(int[][] grid)
    {
 
        // Generate prefix sum of the matrix
        int prefsum[][] = new int[n][m];
        for (int i = 0; i < n; i++) {
            for (int x = 0; x < m; x++) {
                if (x == 0)
                    prefsum[i][x] = grid[i][x];
                else
                    prefsum[i][x]
                        = prefsum[i][x - 1] + grid[i][x];
            }
        }
 
        int dp[][] = new int[n][X + 1];
 
        // Initialize dp[][]
        for (int dpp[] : dp)
            Arrays.fill(dpp, Integer.MIN_VALUE);
 
        // Maximum possible sum by selecting
        // 0 elements from the first i rows
        for (int i = 0; i < n; i++)
            dp[i][0] = 0;
 
        // If a single row is present
        for (int i = 1; i <= Math.min(m, X); ++i) {
            dp[0][i] = dp[0][i - 1] + grid[0][i - 1];
        }
 
        for (int i = 1; i < n; ++i) {
            for (int j = 1; j <= X; ++j) {
 
                // If elements from the
                // current row is not selected
                dp[i][j] = dp[i - 1][j];
 
                // Iterate over all possible
                // selections from current row
                for (int x = 1; x <= Math.min(j, m);
                     x++) {
                    dp[i][j]
                        = Math.max(dp[i][j],
                                   dp[i - 1][j - x]
                                       + prefsum[i][x - 1]);
                }
            }
        }
 
        // Return maximum possible sum
        return dp[n - 1][X];
    }
 
    // Driver Code
    public static void main(String[] args)
    {
        n = 4;
        m = 4;
        X = 6;
 
        int grid[][] = { { 3, 2, 6, 1 },
                         { 1, 9, 2, 4 },
                         { 4, 1, 3, 9 },
                         { 3, 8, 2, 1 } };
 
        int ans = maxSum(grid);
 
        System.out.println(ans);
    }
}

chevron_right


Python3

filter_none

edit
close

play_arrow

link
brightness_4
code

# Python3 program to implement
# the above approach
import sys
 
# Function to calculate the maximum
# possible sum by selecting X elements
# from the Matrix
def maxSum(grid):
     
    # Generate prefix sum of the matrix
    prefsum = [[0 for x in range(m)]
                  for y in range(m)]
                   
    for i in range(n):
        for x in range(m):
            if (x == 0):
                prefsum[i][x] = grid[i][x]
            else:
                prefsum[i][x] = (prefsum[i][x - 1] +
                                    grid[i][x])
 
    dp = [[-sys.maxsize - 1 for x in range(X + 1)]
                            for y in range(n)]
 
    # Maximum possible sum by selecting
    # 0 elements from the first i rows
    for i in range(n):
        dp[i][0] = 0
 
    # If a single row is present
    for i in range(1, min(m, X)):
        dp[0][i] = (dp[0][i - 1] +
                  grid[0][i - 1])
 
    for i in range(1, n):
        for j in range(1, X + 1):
 
            # If elements from the
            # current row is not selected
            dp[i][j] = dp[i - 1][j]
 
            # Iterate over all possible
            # selections from current row
            for x in range(1, min(j, m) + 1):
                    dp[i][j] = max(dp[i][j],
                                   dp[i - 1][j - x] +
                              prefsum[i][x - 1])
 
    # Return maximum possible sum
    return dp[n - 1][X]
 
# Driver Code
if __name__ == "__main__":
     
    n = 4
    m = 4
    X = 6
 
    grid = [ [ 3, 2, 6, 1 ],
             [ 1, 9, 2, 4 ],
             [ 4, 1, 3, 9 ],
             [ 3, 8, 2, 1 ] ]
    ans = maxSum(grid)
 
    print(ans)
 
# This code is contributed by chitranayal   

chevron_right


C#

filter_none

edit
close

play_arrow

link
brightness_4
code

// C# program to implement
// the above approach
using System;
 
class GFG{
 
static int n, m, X;
 
// Function to calculate the maximum
// possible sum by selecting X elements
// from the Matrix
public static int maxSum(int[,] grid)
{
     
    // Generate prefix sum of the matrix
    int [,]prefsum = new int[n, m];
     
    for(int i = 0; i < n; i++)
    {
        for(int x = 0; x < m; x++)
        {
            if (x == 0)
                prefsum[i, x] = grid[i, x];
            else
                prefsum[i, x] = prefsum[i, x - 1] +
                                   grid[i, x];
        }
    }
 
    int [,]dp = new int[n, X + 1];
 
    // Initialize [,]dp
    for(int i = 1; i < n; i++)
        for(int j = 1; j <= X; ++j)
            dp[i, j] = int.MinValue;
             
    // Maximum possible sum by selecting
    // 0 elements from the first i rows
    for(int i = 0; i < n; i++)
        dp[i, 0] = 0;
 
    // If a single row is present
    for(int i = 1; i <= Math.Min(m, X); ++i)
    {
        dp[0, i] = dp[0, i - 1] + grid[0, i - 1];
    }
 
    for(int i = 1; i < n; ++i)
    {
        for(int j = 1; j <= X; ++j)
        {
             
            // If elements from the
            // current row is not selected
            dp[i, j] = dp[i - 1, j];
 
            // Iterate over all possible
            // selections from current row
            for(int x = 1; x <= Math.Min(j, m); x++)
            {
                dp[i, j] = Math.Max(dp[i, j],
                                    dp[i - 1, j - x] +
                               prefsum[i, x - 1]);
            }
        }
    }
     
    // Return maximum possible sum
    return dp[n - 1, X];
}
 
// Driver Code
public static void Main(String[] args)
{
    n = 4;
    m = 4;
    X = 6;
 
    int [,]grid = { { 3, 2, 6, 1 },
                    { 1, 9, 2, 4 },
                    { 4, 1, 3, 9 },
                    { 3, 8, 2, 1 } };
 
    int ans = maxSum(grid);
 
    Console.WriteLine(ans);
}
}
 
// This code is contributed by 29AjayKumar

chevron_right


Output: 

28


 

Time Complexity: O(N*M*X)
Auxiliary Space: O(N*M)

Attention reader! Don’t stop learning now. Get hold of all the important DSA concepts with the DSA Self Paced Course at a student-friendly price and become industry ready.




My Personal Notes arrow_drop_up

Recommended Posts:


COMPUTER SCIENCE ENGINEERING FROM DELHI TECHNOLOGICAL UNIVERSITY

If you like GeeksforGeeks and would like to contribute, you can also write an article using contribute.geeksforgeeks.org or mail your article to contribute@geeksforgeeks.org. See your article appearing on the GeeksforGeeks main page and help other Geeks.

Please Improve this article if you find anything incorrect by clicking on the "Improve Article" button below.