Count of replacements required to make the sum of all Pairs of given type from the Array equal

Given an integer array arr of length N and an integer K, the task is to find the number of array elements to be replaced by a value from the range [1, K] such that each pair (arr[i], arr[N – 1 – i] have equal sum.

Examples :

Input: arr[] = {1, 2, 2, 1}, K = 3 
Output: 1 
Explanation:
Replace arr[0] with 3, so that the array becomes {3, 2, 2, 1}. 
Now, it can be seen that arr[0] + arr[3] = arr[1] + arr[2] = 4.

Input: arr[] = {1, 2, 1, 2, 1, 2} 
Output: 0 
Explanation:
No need to make any changes as arr[0] + arr[5] = arr[1] + arr[4] = arr[2] + arr[3] = 3

Naive Approach :
The simplest approach to this problem could be like iterate for all possible values of X, which can be any number in the range 1 to 2*K, and find the number of operation require to achieve pair sum equal to X. And finally return the minimum numbers of replacement from all operations. 

Time Complexity : O (N * K)
Auxiliary Space : O (1)

Efficient Approach: The above approach can be optimized by the following observations:

• X can clearly take values in the range [2, 2 * K].
• Considering pairs arr[i] and arr[N – i – 1], it can be observed that the sum of any pair can be made equal to X in 0, 1 or 2 replacements.
• Let the maximum of these two numbers be mx and minimum be mn. In one replacement, the sum of this pair will be in the range [mn + 1, mx + K].
• The pair for which mx is less than ( X – K) and mn is greater than or equal to X will need 2 replacements.
• Simply insert mx and mn for all pairs in two different vectors, sort them and apply binary search to find the number of pairs for which 2 replacements are required.
• Find the pairs which do not need any replacement by using Map by storing the frequencies of each sum.
• Finally number of pairs which require one replacement can be calculated by the equation (N / 2 – (pairs which need 2 replacements + pairs which need no replacement) ).

Follow these steps to find the solve the problem: 

1. Find maximum and minimum for all pairs arr [ i ], arr [ N – i – 1 ] and store them in max_values and min_values vectors respectively. Also store the frequencies of sum of all such pairs in a map.
2. Sort the vectors max_values and min_values.
3. Iterate from X = 2 to X = 2 * K, and for every value of X find the total number of replacements as described in the approach above. Update the answer as:

 answer = min ( answer, 2 * (number of pairs for which we need to make 2 replacements) + (number of pairs for which we need to make one replacement) ).

Illustration :
arr[] = { 1, 2, 2, 1 }, K = 3
max_values = {1, 2} 
min_values = {1, 2} 
map = {{2, 1}, {4, 1}}
At X = 4, minimum replacement is required, that is, only 1 replacement, either conversion arr[0] to 3 or arr[3] to 3, is required to make the sum of all pairs equal to 4.

Below is the implementation of the above approach: 

C++

// C++ Program to implement the
// above approach
#include <bits/stdc++.h>
#define int long long int
using namespace std;

const int inf = 1e18;

// Function to find the minimum
// replacements required
int minimumReplacement(int* arr, int N,
                    int K)
{
    int ans = inf;

    // Stores the maximum and minimum
    // values for every pair of
    // the form arr[i], arr[n-i-1]
    vector<int> max_values;
    vector<int> min_values;

    // Map for storing frequencies
    // of every sum formed by pairs
    map<int, int> sum_equal_to_x;

    for (int i = 0; i < N / 2; i++) {

        // Minimum element in the pair
        int mn = min(arr[i], arr[N - i - 1]);

        // Maximum element in the pair
        int mx = max(arr[i], arr[N - i - 1]);

        // Incrementing the frequency of
        // sum encountered
        sum_equal_to_x[arr[i]
                    + arr[N - i - 1]]++;

        // Insert minimum and maximum values
        min_values.push_back(mn);
        max_values.push_back(mx);
    }

    // Sorting the vectors
    sort(max_values.begin(),
        max_values.end());

    sort(min_values.begin(),
        min_values.end());

    // Iterate over all possible values of x
    for (int x = 2; x <= 2 * K; x++) {

        // Count of pairs for which x > x + k
        int mp1
            = lower_bound(max_values.begin(),
                        max_values.end(), x - K)
            - max_values.begin();

        // Count of pairs for which x < mn + 1
        int mp2
            = lower_bound(min_values.begin(),
                        min_values.end(), x)
            - min_values.begin();

        // Count of pairs requiring 2 replacements
        int rep2 = mp1 + (N / 2 - mp2);

        // Count of pairs requiring no replacements
        int rep0 = sum_equal_to_x[x];

        // Count of pairs requiring 1 replacement
        int rep1 = (N / 2 - rep2 - rep0);

        // Update the answer
        ans = min(ans, rep2 * 2 + rep1);
    }

    // Return the answer
    return ans;
}

// Driver Code
int32_t main()
{
    int N = 4;
    int K = 3;
    int arr[] = { 1, 2, 2, 1 };

    cout << minimumReplacement(arr, N, K);
    return 0;
}

Java

// Java program implement
// the above approach
import java.util.*;
import java.io.*;

class GFG{

static int inf = (int)1e18;

// Function to find the minimum
// replacements required
static int minimumReplacement(int[] arr, int N,
                                        int K)
{
    int ans = inf;

    // Stores the maximum and minimum
    // values for every pair of
    // the form arr[i], arr[n-i-1]
    ArrayList<Integer> max_values = new ArrayList<>();
    ArrayList<Integer> min_values = new ArrayList<>();

    // Map for storing frequencies
    // of every sum formed by pairs
    Map<Integer,
        Integer> sum_equal_to_x = new HashMap<>();

    for(int i = 0; i < N / 2; i++)
    {

        // Minimum element in the pair
        int mn = Math.min(arr[i], arr[N - i - 1]);

        // Maximum element in the pair
        int mx = Math.max(arr[i], arr[N - i - 1]);

        // Incrementing the frequency of
        // sum encountered
        sum_equal_to_x.put(arr[i] +
                        arr[N - i - 1],
        sum_equal_to_x.getOrDefault(arr[i] +
                                    arr[N - i - 1],
                                    0) + 1);

        // Insert minimum and maximum values
        min_values.add(mn);
        max_values.add(mx);
    }

    // Sorting the vectors
    Collections.sort(max_values);

    Collections.sort(min_values);

    // Iterate over all possible values of x
    for(int x = 2; x <= 2 * K; x++)
    {

        // Count of pairs for which x > x + k
        int mp1 = max_values.indexOf(x - K);

        // Count of pairs for which x < mn + 1
        int mp2 = min_values.indexOf(x);

        // Count of pairs requiring 2 replacements
        int rep2 = mp1 + (N / 2 - mp2);

        // Count of pairs requiring no replacements
        int rep0 = sum_equal_to_x.getOrDefault(x, -1);

        // Count of pairs requiring 1 replacement
        int rep1 = (N / 2 - rep2 - rep0);

        // Update the answer
        ans = Math.min(ans, rep2 * 2 + rep1);
    }

    // Return the answer
    return ans;
}

// Driver Code
public static void main (String[] args)
{
    int N = 4;
    int K = 3;
    int arr[] = { 1, 2, 2, 1 };

    System.out.print(minimumReplacement(arr, N, K));
}
}

// This code is contributed by offbeat

Python3

# Python3 program to implement the
# above approach
inf = 10**18

def firstOccurrence(numbers, length,
                   searchnum):

    answer = -1
    start = 0
    end = length - 1

    while start <= end:
        middle = (start + end) // 2

        if numbers[middle] == searchnum:
            answer = middle
            end = middle - 1
        elif numbers[middle] > searchnum:
            end = middle - 1
        else:
            start = middle + 1

    return answer

# Function to find the minimum
# replacements required
def minimumReplacement(arr, N,  K):

    ans = inf

    # Stores the maximum and minimum
    # values for every pair of
    # the form arr[i], arr[n-i-1]
    max_values = []
    min_values = []

    # Map for storing frequencies
    # of every sum formed by pairs
    sum_equal_to_x = dict()

    for i in range(N // 2):

        # Minimum element in the pair
        mn = min(arr[i], arr[N - i - 1])

        # Maximum element in the pair
        mx = max(arr[i], arr[N - i - 1])

        # Incrementing the frequency of
        # sum encountered
        if(arr[i] + arr[N - i - 1] not in sum_equal_to_x):
            sum_equal_to_x[arr[i] + arr[N - i - 1]] = 0

        sum_equal_to_x[arr[i] + arr[N - i - 1]] += 1

        # Insert minimum and maximum values
        min_values.append(mn)
        max_values.append(mx)

    max_values.sort()
    min_values.sort()

    # Iterate over all possible values of x
    for x in range(2, 2 * K + 1):

        # Count of pairs for which x > x + k
        mp1 = firstOccurrence(
            max_values, len(max_values), x - K)

        # Count of pairs for which x < mn + 1
        mp2 = firstOccurrence(
            min_values, len(min_values), x)

        # Count of pairs requiring 2 replacements
        rep2 = mp1 + (N // 2 - mp2)

        # Count of pairs requiring no replacements
        if x in sum_equal_to_x:
            rep0 = sum_equal_to_x[x]
        else:
            rep0 = 0

        # Count of pairs requiring 1 replacement
        rep1 = (N // 2 - rep2 - rep0)

        # Update the answer
        ans = min(ans, rep2 * 2 + rep1)

    # Return the answer
    return ans

# Driver Code
if __name__=='__main__':

    N = 4
    K = 3
    arr = [ 1, 2, 2, 1 ]

    print(minimumReplacement(arr, N, K))

# This code is contributed by pratham76

C#

// C# program implement
// the above approach
using System;
using System.Collections;
using System.Collections.Generic;

class GFG{

// Function to find the minimum
// replacements required
static int minimumReplacement(int[] arr, int N,
                                         int K)
{
    int ans = int.MaxValue;

    // Stores the maximum and minimum
    // values for every pair of
    // the form arr[i], arr[n-i-1]
    ArrayList max_values = new ArrayList();
    ArrayList min_values = new ArrayList();

    // Map for storing frequencies
    // of every sum formed by pairs
    Dictionary<int,
               int> sum_equal_to_x = new Dictionary<int,
                                                    int>();

    for(int i = 0; i < N / 2; i++)
    {

        // Minimum element in the pair
        int mn = Math.Min(arr[i], arr[N - i - 1]);

        // Maximum element in the pair
        int mx = Math.Max(arr[i], arr[N - i - 1]);

        // Incrementing the frequency of
        // sum encountered
        sum_equal_to_x[arr[i] + arr[N - i - 1]] =
        sum_equal_to_x.GetValueOrDefault(arr[i] +
                             arr[N - i - 1], 0) + 1;

        // Insert minimum and maximum values
        min_values.Add(mn);
        max_values.Add(mx);
    }

    // Sorting the vectors
    max_values.Sort();

    min_values.Sort();

    // Iterate over all possible values of x
    for(int x = 2; x <= 2 * K; x++)
    {

        // Count of pairs for which x > x + k
        int mp1 = max_values.IndexOf(x - K);

        // Count of pairs for which x < mn + 1
        int mp2 = min_values.IndexOf(x);

        // Count of pairs requiring 2 replacements
        int rep2 = mp1 + (N / 2 - mp2);

        // Count of pairs requiring no replacements
        int rep0 = sum_equal_to_x.GetValueOrDefault(
                   x, -1);

        // Count of pairs requiring 1 replacement
        int rep1 = (N / 2 - rep2 - rep0);

        // Update the answer
        ans = Math.Min(ans, rep2 * 2 + rep1);
    }

    // Return the answer
    return ans;
}

// Driver Code
public static void Main(string[] args)
{
    int N = 4;
    int K = 3;
    int []arr = { 1, 2, 2, 1 };

    Console.Write(minimumReplacement(arr, N, K));
}
}

// This code is contributed by rutvik_56

Javascript

<script>
// JavaScript program to implement the
// above approach
const inf = 10**18

function firstOccurrence(numbers, length, searchnum)
{

    let answer = -1
    let start = 0
    let end = length - 1

    while(start <= end){
        let middle = Math.floor((start + end)/2)

        if(numbers[middle] == searchnum){
            answer = middle
            end = middle - 1
        }
        else if(numbers[middle] > searchnum)
            end = middle - 1
        else
            start = middle + 1
    }

    return answer
}

// Function to find the minimum
// replacements required
function minimumReplacement(arr, N, K){

    let ans = inf

    // Stores the maximum and minimum
    // values for every pair of
    // the form arr[i], arr[n-i-1]
    let max_values = []
    let min_values = []

    // Map for storing frequencies
    // of every sum formed by pairs
    let sum_equal_to_x = new Map()

    for(let i=0;i<Math.floor(N / 2);i++){

        // Minimum element in the pair
        let mn = Math.min(arr[i], arr[N - i - 1])

        // Maximum element in the pair
        let mx = Math.max(arr[i], arr[N - i - 1])

        // Incrementing the frequency of
        // sum encountered
        if(sum_equal_to_x.has(arr[i] + arr[N - i - 1])==false)
            sum_equal_to_x.set(arr[i] + arr[N - i - 1],0)

        sum_equal_to_x.set(arr[i] + arr[N - i - 1],
        sum_equal_to_x.get(arr[i] + arr[N - i -1])+1)

        // Insert minimum and maximum values
        min_values.push(mn)
        max_values.push(mx)
    }

    max_values.sort()
    min_values.sort()

    // Iterate over all possible values of x
    for(let x=2;x<2 * K + 1;x++){

        // Count of pairs for which x > x + k
        let mp1 = firstOccurrence(max_values, max_values.length, x - K)

        // Count of pairs for which x < mn + 1
        let mp2 = firstOccurrence(min_values, min_values.length, x)

        // Count of pairs requiring 2 replacements
        rep2 = mp1 + (Math.floor(N / 2) - mp2)

        // Count of pairs requiring no replacements
        if(sum_equal_to_x.has(x)){
            rep0 = sum_equal_to_x.get(x)
        }
        else
            rep0 = 0

        // Count of pairs requiring 1 replacement
        rep1 = Math.floor(N /2) - rep2 - rep0

        // Update the answer
        ans = Math.min(ans, rep2 * 2 + rep1)
    }

    // Return the answer
    return ans
}

// Driver Code

let N = 4
let K = 3
let arr = [ 1, 2, 2, 1 ]

document.write(minimumReplacement(arr, N, K))

// This code is contributed by shinjanpatra

</script>

Output

1

Time Complexity: O(NlogN)
Auxiliary Space: O(N)