Tutorialspoint.dev

Maximum product of an increasing subsequence

Given an array of numbers, find the maximum product formed by multiplying numbers of an increasing subsequence of that array.

Note: A single number is supposed to be an increasing subsequence of size 1.

Examples:

Input : arr[] = { 3, 100, 4, 5, 150, 6 }
Output : 45000
Maximum product is 45000 formed by the 
increasing subsequence 3, 100, 150. Note
that the longest increasing subsequence 
is different {3, 4, 5, 6}

Input : arr[] = { 10, 22, 9, 33, 21, 50, 41, 60 }
Output : 21780000
Maximum product is 21780000 formed by the 
increasing subsequence 10, 22, 33, 50, 60.
          



Prerequisite : Longest Increasing Subsequence

Approach: Use a dynamic approach to maintain a table mpis[]. The value of mpis[i] stores product maximum product increasing subsequence ending with arr[i]. Initially all the values of increasing subsequence table are initialized to arr[i]. We use recursive approach similar to LIS problem to find the result.

C++

/* Dynamic programming C++ implementation of maximum 
   product of an increasing subsequence */
#include <bits/stdc++.h>
#define ll long long int
using namespace std;
  
// Returns product of maximum product increasing
// subsequence.
ll lis(ll arr[], ll n)
{
    ll mpis[n];
  
    /* Initialize MPIS values */
    for (int i = 0; i < n; i++)
        mpis[i] = arr[i];
  
    /* Compute optimized MPIS values considering
       every element as ending element of sequence */
    for (int i = 1; i < n; i++)
        for (int j = 0; j < i; j++)
            if (arr[i] > arr[j] && mpis[i] < (mpis[j] * arr[i]))
                mpis[i] = mpis[j] * arr[i];
  
    /* Pick maximum of all product values */
    return *max_element(mpis, mpis + n);
}
  
/* Driver program to test above function */
int main()
{
    ll arr[] = { 3, 100, 4, 5, 150, 6 };
    ll n = sizeof(arr) / sizeof(arr[0]);
    printf("%lld", lis(arr, n));
    return 0;
}

Java

/* Dynamic programming Java implementation 
of maximum product of an increasing 
subsequence */
import java.util.Arrays;
import java.util.Collections;
  
class GFG {
  
    // Returns product of maximum product
    // increasing subsequence.
    static int lis(int[] arr, int n)
    {
        int[] mpis = new int[n];
        int max = Integer.MIN_VALUE;
          
        /* Initialize MPIS values */
        for (int i = 0; i < n; i++)
            mpis[i] = arr[i];
  
        /* Compute optimized MPIS values 
        considering every element as ending
        element of sequence */
        for (int i = 1; i < n; i++)
            for (int j = 0; j < i; j++)
                if (arr[i] > arr[j] && mpis[i] 
                         < (mpis[j] * arr[i]))
                    mpis[i] = mpis[j] * arr[i];
  
        /* Pick maximum of all product values 
        using for loop*/
        for (int k = 0; k < mpis.length; k++)
        {
            if (mpis[k] > max) {
                max = mpis[k];
            }
        }
          
        return max;
    }
  
    // Driver program to test above function
    static public void main(String[] args)
    {
  
        int[] arr = { 3, 100, 4, 5, 150, 6 };
        int n = arr.length;
  
        System.out.println(lis(arr, n));
    }
}
  
// This code is contributed by parashar.

Python3

# Dynamic programming Python3 implementation
# of maximum product of an increasing
# subsequence 
  
# Returns product of maximum product
# increasing subsequence.
def lis (arr, n ):
    mpis =[0] * (n)
      
    # Initialize MPIS values
    for i in range(n):
        mpis[i] = arr[i]
      
    # Compute optimized MPIS values
    # considering every element as 
    # ending element of sequence
    for i in range(1, n):
        for j in range(i):
            if (arr[i] > arr[j] and
                    mpis[i] < (mpis[j] * arr[i])):
                        mpis[i] = mpis[j] * arr[i]
      
    # Pick maximum of all product values 
    return max(mpis)
  
# Driver code to test above function
arr = [3, 100, 4, 5, 150, 6]
n = len(arr)
print( lis(arr, n))
  
# This code is contributed by "Sharad_Bhardwaj".

C#

/* Dynamic programming C# implementation 
of maximum product of an increasing 
subsequence */
using System;
using System.Linq;
  
public class GFG {
  
    // Returns product of maximum product
    // increasing subsequence.
    static long lis(long[] arr, long n)
    {
        long[] mpis = new long[n];
  
        /* Initialize MPIS values */
        for (int i = 0; i < n; i++)
            mpis[i] = arr[i];
  
        /* Compute optimized MPIS values considering
        every element as ending element of sequence */
        for (int i = 1; i < n; i++)
            for (int j = 0; j < i; j++)
                if (arr[i] > arr[j] && mpis[i] < (mpis[j] * arr[i]))
                    mpis[i] = mpis[j] * arr[i];
  
        /* Pick maximum of all product values */
        return mpis.Max();
    }
  
    /* Driver program to test above function */
    static public void Main()
    {
  
        long[] arr = { 3, 100, 4, 5, 150, 6 };
        long n = arr.Length;
  
        Console.WriteLine(lis(arr, n));
    }
}
  
// This code is contributed by vt_m.

PHP

<?PHP 
  
/* Dynamic programming PHP implementation of maximum 
   product of an increasing subsequence */
   
// Returns product of maximum product increasing
// subsequence.
function lis(&$arr$n)
{
    $mpis = array_fill(0,$n, NULL);
   
    /* Initialize MPIS values */
    for ($i = 0; $i < $n; $i++)
        $mpis[$i] = $arr[$i];
   
    /* Compute optimized MPIS values considering
       every element as ending element of sequence */
    for ($i = 1; $i < $n; $i++)
        for ($j = 0; $j < $i; $j++)
            if ($arr[$i] > $arr[$j] && $mpis[$i] < ($mpis[$j] * $arr[$i]))
                $mpis[$i] = $mpis[$j] * $arr[$i];
   
    /* Pick maximum of all product values */
    return max($mpis);
}
   
/* Driver program to test above function */
  
    $arr = array ( 3, 100, 4, 5, 150, 6 );
    $n = sizeof($arr) / sizeof($arr[0]);
    echo lis($arr, $n);
    return 0;
?>


Output:

 45000

Time Complexity: O(n^2)
Auxiliary Space : O(n)



This article is attributed to GeeksforGeeks.org

leave a comment

code

0 Comments

load comments

Subscribe to Our Newsletter