Given N elements, write a program that prints the length of the longest increasing subsequence whose adjacent element difference is one.

Examples:

Input : a[] = {3, 10, 3, 11, 4, 5, 6, 7, 8, 12}

Output : 6

Explanation: 3, 4, 5, 6, 7, 8 is the longest increasing subsequence whose adjacent element differs by one.Input : a[] = {6, 7, 8, 3, 4, 5, 9, 10}

Output : 5

Explanation: 6, 7, 8, 9, 10 is the longest increasing subsequence

**Naive Approach: ** A normal approach will be to iterate for every element and find out the longest increasing subsequence. For any particular element, find the length of the subsequence starting from that element. Print the longest length of the subsequence thus formed. The time complexity of this approach will be O(n^{2}).

**Dynamic Programming Approach: ** Let DP[i] store the length of the longest subsequence which ends with A[i]. For every A[i], if A[i]-1 is present in the array before i-th index, then A[i] will add to the increasing subsequence which has A[i]-1. Hence **DP[i] = DP[ index(A[i]-1) ] + 1**. If A[i]-1 is not present in the array before i-th index, then DP[i]=1 since the A[i] element forms a subsequence which starts with A[i]. Hence the relation for DP[i] is:

If A[i]-1 is present before i-th index:

DP[i] = DP[ index(A[i]-1) ] + 1 else:

DP[i] = 1

Given below is the illustration of the above approach:

// CPP program to find length of the // longest increasing subsequence // whose adjacent element differ by 1 #include <bits/stdc++.h> using namespace std; // function that returns the length of the // longest increasing subsequence // whose adjacent element differ by 1 int longestSubsequence(int a[], int n) { // stores the index of elements unordered_map<int, int> mp; // stores the length of the longest // subsequence that ends with a[i] int dp[n]; memset(dp, 0, sizeof(dp)); int maximum = INT_MIN; // iterate for all element for (int i = 0; i < n; i++) { // if a[i]-1 is present before i-th index if (mp.find(a[i] - 1) != mp.end()) { // last index of a[i]-1 int lastIndex = mp[a[i] - 1] - 1; // relation dp[i] = 1 + dp[lastIndex]; } else dp[i] = 1; // stores the index as 1-index as we need to // check for occurrence, hence 0-th index // will not be possible to check mp[a[i]] = i + 1; // stores the longest length maximum = max(maximum, dp[i]); } return maximum; } // Driver Code int main() { int a[] = { 3, 10, 3, 11, 4, 5, 6, 7, 8, 12 }; int n = sizeof(a) / sizeof(a[0]); cout << longestSubsequence(a, n); return 0; }

**Output:**

6

Time Complexity : O(n)

Auxiliary Space : O(n)

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 write comments if you find anything incorrect, or you want to share more information about the topic discussed above.