Given a binary array arr[], the task is to find the maximum possible length of non-decreasing subsequence that can be generated by reversing a subarray at most once.
Examples:
Input: arr[] = {0, 1, 0, 1}
Output: 4
Explanation:
After reversing the subarray from index [2, 3], the array modifies to {0, 0, 1, 1}.
Hence, the longest non-decreasing subsequence is {0, 0, 1, 1}.
Input: arr[] = {0, 1, 1, 1, 0, 0, 1, 1, 0}
Output: 9
Explanation:
After reversing the subarray from index [2, 6], the array modifies to {0, 0, 0, 1, 1, 1, 1, 1, 0}.
Hence, the longest non-decreasing subsequence is {0, 0, 0, 1, 1, 1, 1, 1}.
Naive Approach: The simplest approach to solve the problem is to reverse each possible subarray in the given array, and find the longest non-decreasing subsequence possible from the array after reversing the subarray.
Time Complexity: O(N3)
Auxiliary Space: O(N)
Efficient Approach: The idea is to use Dynamic Programming to solve the problem. Follow the steps below:
- Since the array is a binary array the idea is to find the longest subsequence among the subsequences of the forms {0....0}, {0...1...}, {0..1..0...}, 0..1..0..1.
- Initialize a dynamic programming table as dp[][] which stores the following:
dp[i][0] : Stores the length of the longest subsequence (0..) from a[0 to i].
dp[i][1] : Stores the length of the longest subsequence (0..1..) from a[0 to i].
dp[i][2] : Stores the length of the longest subsequence (0..1..0..) from a[0 to i].
dp[i][3] : Stores the length of the longest subsequence (0..1..0..1..) from a[0 to i].
- Therefore, the answer is the longest subsequence or the maximum of all the 4 given possibilities ( dp[n-1][0], d[n-1][1], dp[n-1][2], dp[n-1][3] ).
Below is the implementation of the above approach:
// C++ program to implement
// the above approach
#include<bits/stdc++.h>
using namespace std;
// Function to find the maximum length
// non decreasing subarray by reversing
// at most one subarray
void main_fun(int arr[], int n)
{
// dp[i][j] be the longest
// subsequence of a[0...i]
// with first j parts
int dp[4][n];
memset(dp, 0, sizeof(dp[0][0] * 4 * n));
if (arr[0] == 0)
dp[0][0] = 1;
else
dp[1][0] = 1;
// Maximum length sub-sequence
// of (0..)
for(int i = 1; i < n; i++)
{
if (arr[i] == 0)
dp[0][i] = dp[0][i - 1] + 1;
else
dp[0][i] = dp[0][i - 1];
}
// Maximum length sub-sequence
// of (0..1..)
for(int i = 1; i < n; i++)
{
if (arr[i] == 1)
dp[1][i] = max(dp[1][i - 1] + 1,
dp[0][i - 1] + 1);
else
dp[1][i] = dp[1][i - 1];
}
// Maximum length sub-sequence
// of (0..1..0..)
for(int i = 1; i < n; i++)
{
if (arr[i] == 0)
{
dp[2][i] = max(dp[2][i - 1] + 1,
max(dp[1][i - 1] + 1,
dp[0][i - 1] + 1));
}
else
dp[2][i] = dp[2][i - 1];
}
// Maximum length sub-sequence
// of (0..1..0..1..)
for(int i = 1; i < n; i++)
{
if (arr[i] == 1)
{
dp[3][i] = max(dp[3][i - 1] + 1,
max(dp[2][i - 1] + 1,
max(dp[1][i - 1] + 1,
dp[0][i - 1] + 1)));
}
else
dp[3][i] = dp[3][i - 1];
}
// Find the max length subsequence
int ans = max(dp[2][n - 1], max(dp[1][n - 1],
max(dp[0][n - 1], dp[3][n - 1])));
// Print the answer
cout << (ans);
}
// Driver Code
int main()
{
int n = 4;
int arr[] = {0, 1, 0, 1};
main_fun(arr, n);
return 0;
}
// This code is contributed by chitranayal
// Java program to implement
// the above approach
import java.util.*;
class GFG{
// Function to find the maximum length
// non decreasing subarray by reversing
// at most one subarray
static void main_fun(int arr[], int n)
{
// dp[i][j] be the longest
// subsequence of a[0...i]
// with first j parts
int[][] dp = new int[4][n];
if (arr[0] == 0)
dp[0][0] = 1;
else
dp[1][0] = 1;
// Maximum length sub-sequence
// of (0..)
for(int i = 1; i < n; i++)
{
if (arr[i] == 0)
dp[0][i] = dp[0][i - 1] + 1;
else
dp[0][i] = dp[0][i - 1];
}
// Maximum length sub-sequence
// of (0..1..)
for(int i = 1; i < n; i++)
{
if (arr[i] == 1)
dp[1][i] = Math.max(dp[1][i - 1] + 1,
dp[0][i - 1] + 1);
else
dp[1][i] = dp[1][i - 1];
}
// Maximum length sub-sequence
// of (0..1..0..)
for(int i = 1; i < n; i++)
{
if (arr[i] == 0)
{
dp[2][i] = Math.max(dp[2][i - 1] + 1,
Math.max(dp[1][i - 1] + 1,
dp[0][i - 1] + 1));
}
else
dp[2][i] = dp[2][i - 1];
}
// Maximum length sub-sequence
// of (0..1..0..1..)
for(int i = 1; i < n; i++)
{
if (arr[i] == 1)
{
dp[3][i] = Math.max(dp[3][i - 1] + 1,
Math.max(dp[2][i - 1] + 1,
Math.max(dp[1][i - 1] + 1,
dp[0][i - 1] + 1)));
}
else
dp[3][i] = dp[3][i - 1];
}
// Find the max length subsequence
int ans = Math.max(dp[2][n - 1],
Math.max(dp[1][n - 1],
Math.max(dp[0][n - 1],
dp[3][n - 1])));
// Print the answer
System.out.print(ans);
}
// Driver code
public static void main (String[] args)
{
int n = 4;
int arr[] = { 0, 1, 0, 1 };
main_fun(arr, n);
}
}
// This code is contributed by offbeat
# Python3 program to implement
# the above approach
import sys
# Function to find the maximum length
# non decreasing subarray by reversing
# at most one subarray
def main(arr, n):
# dp[i][j] be the longest
# subsequence of a[0...i]
# with first j parts
dp = [[0 for x in range(n)] for y in range(4)]
if arr[0] == 0:
dp[0][0] = 1
else:
dp[1][0] = 1
# Maximum length sub-sequence
# of (0..)
for i in range(1, n):
if arr[i] == 0:
dp[0][i] = dp[0][i-1] + 1
else:
dp[0][i] = dp[0][i-1]
# Maximum length sub-sequence
# of (0..1..)
for i in range(1, n):
if arr[i] == 1:
dp[1][i] = max(dp[1][i-1] + 1, dp[0][i-1] + 1)
else:
dp[1][i] = dp[1][i-1]
# Maximum length sub-sequence
# of (0..1..0..)
for i in range(1, n):
if arr[i] == 0:
dp[2][i] = max([dp[2][i-1] + 1,
dp[1][i-1] + 1,
dp[0][i-1] + 1])
else:
dp[2][i] = dp[2][i-1]
# Maximum length sub-sequence
# of (0..1..0..1..)
for i in range(1, n):
if arr[i] == 1:
dp[3][i] = max([dp[3][i-1] + 1,
dp[2][i-1] + 1,
dp[1][i-1] + 1,
dp[0][i-1] + 1])
else:
dp[3][i] = dp[3][i-1]
# Find the max length subsequence
ans = max([dp[2][n-1], dp[1][n-1],
dp[0][n-1], dp[3][n-1]])
# Print the answer
print(ans)
# Driver Code
if __name__ == "__main__":
n = 4
arr = [0, 1, 0, 1]
main(arr, n)
// C# program to implement
// the above approach
using System;
class GFG{
// Function to find the maximum length
// non decreasing subarray by reversing
// at most one subarray
static void main_fun(int []arr, int n)
{
// dp[i,j] be the longest
// subsequence of a[0...i]
// with first j parts
int[,] dp = new int[4, n];
if (arr[0] == 0)
dp[0, 0] = 1;
else
dp[1, 0] = 1;
// Maximum length sub-sequence
// of (0..)
for(int i = 1; i < n; i++)
{
if (arr[i] == 0)
dp[0, i] = dp[0, i - 1] + 1;
else
dp[0, i] = dp[0, i - 1];
}
// Maximum length sub-sequence
// of (0..1..)
for(int i = 1; i < n; i++)
{
if (arr[i] == 1)
dp[1, i] = Math.Max(dp[1, i - 1] + 1,
dp[0, i - 1] + 1);
else
dp[1, i] = dp[1, i - 1];
}
// Maximum length sub-sequence
// of (0..1..0..)
for(int i = 1; i < n; i++)
{
if (arr[i] == 0)
{
dp[2, i] = Math.Max(dp[2, i - 1] + 1,
Math.Max(dp[1, i - 1] + 1,
dp[0, i - 1] + 1));
}
else
dp[2, i] = dp[2, i - 1];
}
// Maximum length sub-sequence
// of (0..1..0..1..)
for(int i = 1; i < n; i++)
{
if (arr[i] == 1)
{
dp[3, i] = Math.Max(dp[3, i - 1] + 1,
Math.Max(dp[2, i - 1] + 1,
Math.Max(dp[1, i - 1] + 1,
dp[0, i - 1] + 1)));
}
else
dp[3, i] = dp[3, i - 1];
}
// Find the max length subsequence
int ans = Math.Max(dp[2, n - 1],
Math.Max(dp[1, n - 1],
Math.Max(dp[0, n - 1],
dp[3, n - 1])));
// Print the answer
Console.Write(ans);
}
// Driver code
public static void Main(String[] args)
{
int n = 4;
int []arr = { 0, 1, 0, 1 };
main_fun(arr, n);
}
}
// This code is contributed by Amit Katiyar
<script>
// JavaScript program to implement
// the above approach
// Function to find the maximum length
// non decreasing subarray by reversing
// at most one subarray
function main_fun(arr, n)
{
// dp[i][j] be the longest
// subsequence of a[0...i]
// with first j parts
var dp = Array.from(Array(4), ()=>Array(n).fill(0));
if (arr[0] == 0)
dp[0][0] = 1;
else
dp[1][0] = 1;
// Maximum length sub-sequence
// of (0..)
for(var i = 1; i < n; i++)
{
if (arr[i] == 0)
dp[0][i] = dp[0][i - 1] + 1;
else
dp[0][i] = dp[0][i - 1];
}
// Maximum length sub-sequence
// of (0..1..)
for(var i = 1; i < n; i++)
{
if (arr[i] == 1)
dp[1][i] = Math.max(dp[1][i - 1] + 1,
dp[0][i - 1] + 1);
else
dp[1][i] = dp[1][i - 1];
}
// Maximum length sub-sequence
// of (0..1..0..)
for(var i = 1; i < n; i++)
{
if (arr[i] == 0)
{
dp[2][i] = Math.max(dp[2][i - 1] + 1,
Math.max(dp[1][i - 1] + 1,
dp[0][i - 1] + 1));
}
else
dp[2][i] = dp[2][i - 1];
}
// Maximum length sub-sequence
// of (0..1..0..1..)
for(var i = 1; i < n; i++)
{
if (arr[i] == 1)
{
dp[3][i] = Math.max(dp[3][i - 1] + 1,
Math.max(dp[2][i - 1] + 1,
Math.max(dp[1][i - 1] + 1,
dp[0][i - 1] + 1)));
}
else
dp[3][i] = dp[3][i - 1];
}
// Find the max length subsequence
var ans = Math.max(dp[2][n - 1], Math.max(dp[1][n - 1],
Math.max(dp[0][n - 1], dp[3][n - 1])));
// Print the answer
document.write(ans);
}
// Driver Code
var n = 4;
var arr = [0, 1, 0, 1];
main_fun(arr, n);
</script>
Output:
4
Time Complexity: O(N)
Auxiliary Space: O(N) <= O(4*N)