Open In App

Count unique subsequences of length K

Last Updated : 20 Dec, 2022
Improve
Improve
Like Article
Like
Save
Share
Report

Given an array of N numbers and an integer K. The task is to print the number of unique subsequences possible of length K. 

Examples:

Input : a[] = {1, 2, 3, 4}, k = 3 
Output : 4. 
Unique Subsequences are: 
{1, 2, 3}, {1, 2, 4}, {1, 3, 4}, {2, 3, 4}

Input: a[] = {1, 1, 1, 2, 2, 2 }, k = 3
Output : 4 
Unique Subsequences are 
{1, 1, 1}, {1, 1, 2}, {1, 2, 2}, {2, 2, 2} 

Approach: 

There is a well-known formula for how many subsequences of fixed length K can be chosen from N unique objects. But the problem here has several differences. One among them is the order in subsequences is important and must be preserved as in the original sequence. For such a problem there can be no ready combinatorics formula because the results depend on the order of the original array. 

The main idea is to deal recurrently by the length of the subsequence. On each recurrent step, move from the end to the beginning and count the unique combinations using the count of shorter unique combinations from the previous step. More strictly on every step j, we keep an array of length N and every element in the place p means how many unique subsequences with length j we found to the right of the element in place i, including i itself.

Below is the implementation of the above approach.

C++




#include <bits/stdc++.h>
using namespace std;
 
// Function which returns the number of
// unique subsequences of length K
int solution(vector<int>& A, int k)
{
    // size of the vector
    // which is constant
    const int N = A.size();
 
    // base cases
    if (N < k || N < 1 || k < 1)
        return 0;
    if (N == k)
        return 1;
 
    // Prepare vectors for recursion
    vector<int> v1(N, 0);
    vector<int> v2(N, 0);
    vector<int> v3(N, 0);
 
    // initialize separately for k = 1
    // initialize the last element
    v2[N - 1] = 1;
    v3[A[N - 1] - 1] = 1;
 
    // initiate all other elements of k = 1
    for (int i = N - 2; i >= 0; i--) {
 
        // initialize the front element
        // to vector v2
        v2[i] = v2[i + 1];
 
        // if element v[a[i]-1] is 0
        // then increment it in vector v2
        if (v3[A[i] - 1] == 0) {
            v2[i]++;
            v3[A[i] - 1] = 1;
        }
    }
 
    // iterate for all possible values of K
    for (int j = 1; j < k; j++) {
 
        // fill the vectors with 0
        fill(v3.begin(), v3.end(), 0);
 
        // fill(v1.begin(), v1.end(), 0)
        // the last must be 0 as from last no unique
        // subarray can be formed
        v1[N - 1] = 0;
 
        // Iterate for all index from which unique
        // subsequences can be formed
        for (int i = N - 2; i >= 0; i--) {
 
            // add the number of subsequence formed
            // from the next index
            v1[i] = v1[i + 1];
 
            // start with combinations on the
            // next index
            v1[i] = v1[i] + v2[i + 1];
 
            // Remove the elements which have
            // already been counted
            v1[i] = v1[i] - v3[A[i] - 1];
 
            // Update the number used
            v3[A[i] - 1] = v2[i + 1];
        }
 
        // prepare the next iteration
        // by filling v2 in v1
        v2 = v1;
    }
 
    // last answer is stored in v2
    return v2[0];
}
 
// Function to push the vector into an array
// and print all the unique subarrays
void solve(int a[], int n, int k)
{
    vector<int> v;
 
    // fill the vector with a[]
    v.assign(a, a + n);
 
    // Function call to print the count
    // of unique subsequences of size K
    cout << solution(v, k);
}
 
// Driver Code
int main()
{
    int a[] = { 1, 2, 3, 4 };
    int n = sizeof(a) / sizeof(a[0]);
    int k = 3;
    solve(a, n, k);
 
    return 0;
}


Java




import java.util.*;
 
class GFG{
 
// Function which returns the numbe of
// unique subsequences of length K
static int solution(int[] A, int N, int k)
{
 
    // Bases cases
    if (N < k || N < 1 || k < 1)
        return 0;
    if (N == k)
        return 1;
 
    // Prepare arrays for recursion
    int[] v1 = new int[N];
    int[] v2 = new int[N];
    int[] v3 = new int[N];
 
    // Initiate separately for k = 1
    // initiate the last element
    v2[N - 1] = 1;
    v3[A[N - 1] - 1] = 1;
 
    // Initiate all other elements of k = 1
    for(int i = N - 2; i >= 0; i--)
    {
 
        // Initialize the front element
        // to vector v2
        v2[i] = v2[i + 1];
 
        // If element v[a[i]-1] is 0
        // then increment it in vector v2
        if (v3[A[i] - 1] == 0)
        {
            v2[i]++;
            v3[A[i] - 1] = 1;
        }
    }
 
    // Iterate for all possible values of K
    for(int j = 1; j < k; j++)
    {
 
        // Fill the vectors with 0
        Arrays.fill(v3, 0);
 
        // Fill(v1.begin(), v1.end(), 0)
        // the last must be 0 as from last
        // no unique subarray can be formed
        v1[N - 1] = 0;
 
        // Iterate for all index from which
        // unique subsequences can be formed
        for(int i = N - 2; i >= 0; i--)
        {
 
            // Add the number of subsequence
            // formed from the next index
            v1[i] = v1[i + 1];
 
            // Start with combinations on the
            // next index
            v1[i] = v1[i] + v2[i + 1];
 
            // Remove the elements which have
            // already been counted
            v1[i] = v1[i] - v3[A[i] - 1];
 
            // Update the number used
            v3[A[i] - 1] = v2[i + 1];
        }
    }
 
    // Last answer is stored in v2
    return v2[0];
}
 
// Driver Code
public static void main(String[] args)
{
    int a[] = { 1, 2, 3, 4 };
    int n = a.length;
    int k = 3;
     
    System.out.print(solution(a, n, k));
}
}
 
// This code is contributed by amal kumar choubey


Python3




# Function which returns the numbe of
# unique subsequences of length K
def solution( A, k):
 
    # seiz of the vector
    # which does is constant
    N = len(A)
 
    # bases cases
    if (N < k or N < 1 or k < 1):
        return 0
    if (N == k):
        return 1
 
    # Prepare arrays for recursion
    v1 = [0]*(N)
    v2 = [0]*N
    v3 = [0]*N
 
    # initiate separately for k = 1
    # initiate the last element
    v2[N - 1] = 1
    v3[A[N - 1] - 1] = 1
 
    # initiate all other elements of k = 1
    for i in range(N - 2,-1,-1):
 
        # initialize the front element
        # to vector v2
        v2[i] = v2[i + 1]
 
        # if element v[a[i]-1] is 0
        # then increment it in vector v2
        if (v3[A[i] - 1] == 0):
            v2[i] += 1
            v3[A[i] - 1] = 1
         
    # iterate for all possible values of K
    for j in range( 1, k) :
 
        # fill the vectors with 0
        v3 = [0]*N
 
        # fill(v1.begin(), v1.end(), 0)
        # the last must be 0 as from last no unique
        # subarray can be formed
        v1[N - 1] = 0
 
        # Iterate for all index from which unique
        # subsequences can be formed
        for i in range( N - 2, -1, -1) :
 
            # add the number of subsequence formed
            # from the next index
            v1[i] = v1[i + 1]
 
            # start with combinations on the
            # next index
            v1[i] = v1[i] + v2[i + 1]
 
            # Remove the elements which have
            # already been counted
            v1[i] = v1[i] - v3[A[i] - 1]
 
            # Update the number used
            v3[A[i] - 1] = v2[i + 1]
     
        # prepare the next iteration
        # by filling v2 in v1
        for i in range(len(v1)):
            v2[i] = v1[i]
     
    # last answer is stored in v2
    return v2[0]
 
# Function to push the vector into an array
# and print all the unique subarrays
def solve(a, n, k):
 
    # fill the vector with a[]
    v = a
 
    # Function call to print the count
    # of unique subsequences of size K
    print( solution(v, k))
 
# Driver Code
if __name__ == "__main__":
 
    a = [ 1, 2, 3, 4 ]
    n = len(a)
    k = 3
    solve(a, n, k)
 
# This code is contributed by chitranayal


C#




using System;
class GFG{
 
// Function which returns the numbe of
// unique subsequences of length K
static int solution(int[] A, int N, int k)
{
 
    // Bases cases
    if (N < k || N < 1 || k < 1)
        return 0;
    if (N == k)
        return 1;
 
    // Prepare arrays for recursion
    int[] v1 = new int[N];
    int[] v2 = new int[N];
    int[] v3 = new int[N];
 
    // Initiate separately for k = 1
    // initiate the last element
    v2[N - 1] = 1;
    v3[A[N - 1] - 1] = 1;
 
    // Initiate all other elements of k = 1
    for(int i = N - 2; i >= 0; i--)
    {
 
        // Initialize the front element
        // to vector v2
        v2[i] = v2[i + 1];
 
        // If element v[a[i]-1] is 0
        // then increment it in vector v2
        if (v3[A[i] - 1] == 0)
        {
            v2[i]++;
            v3[A[i] - 1] = 1;
        }
    }
 
    // Iterate for all possible values of K
    for(int j = 1; j < k; j++)
    {
 
        // Fill the vectors with 0
        for(int i = 0; i < v3.GetLength(0); i++)
            v3[i] = 0;
 
        // Fill(v1.begin(), v1.end(), 0)
        // the last must be 0 as from last
        // no unique subarray can be formed
        v1[N - 1] = 0;
 
        // Iterate for all index from which
        // unique subsequences can be formed
        for(int i = N - 2; i >= 0; i--)
        {
 
            // Add the number of subsequence
            // formed from the next index
            v1[i] = v1[i + 1];
 
            // Start with combinations on the
            // next index
            v1[i] = v1[i] + v2[i + 1];
 
            // Remove the elements which have
            // already been counted
            v1[i] = v1[i] - v3[A[i] - 1];
 
            // Update the number used
            v3[A[i] - 1] = v2[i + 1];
        }
    }
 
    // Last answer is stored in v2
    return v2[0];
}
 
// Driver Code
public static void Main(String[] args)
{
    int []a = { 1, 2, 3, 4 };
    int n = a.Length;
    int k = 3;
     
    Console.Write(solution(a, n, k));
}
}
 
// This code is contributed by Rohit_ranjan


Javascript




<script>
     
// Function which returns the numbe of
// unique subsequences of length K
function solution(A, N, K)
{
     
    // Bases cases
    if (N < k || N < 1 || k < 1)
        return 0;
    if (N == k)
        return 1;
     
    // Prepare arrays for recursion
    let v1 = new Array(N);
    let v2 = new Array(N);
    let v3 = new Array(N);
     
    for(let i = 0; i < N; i++)
    {
        v1[i] = 0;
        v2[i] = 0;
        v3[i] = 0;
    }
     
    // Initiate separately for k = 1
    // initiate the last element
    v2[N - 1] = 1;
    v3[A[N - 1] - 1] = 1;
     
    // Initiate all other elements of k = 1
    for(let i = N - 2; i >= 0; i--)
    {
         
        // Initialize the front element
        // to vector v2
        v2[i] = v2[i + 1];
     
        // If element v[a[i]-1] is 0
        // then increment it in vector v2
        if (v3[A[i] - 1] == 0)
        {
            v2[i]++;
            v3[A[i] - 1] = 1;
        }
    }
     
    // Iterate for all possible values of K
    for(let j = 1; j < k; j++)
    {
     
        // Fill the vectors with 0
        for(let i = 0; i < v3.length; i++)
        {
            v3[i] = 0;
        }
     
        // Fill(v1.begin(), v1.end(), 0)
        // the last must be 0 as from last
        // no unique subarray can be formed
        v1[N - 1] = 0;
     
        // Iterate for all index from which
        // unique subsequences can be formed
        for(let i = N - 2; i >= 0; i--)
        {
     
            // Add the number of subsequence
            // formed from the next index
            v1[i] = v1[i + 1];
     
            // Start with combinations on the
            // next index
            v1[i] = v1[i] + v2[i + 1];
     
            // Remove the elements which have
            // already been counted
            v1[i] = v1[i] - v3[A[i] - 1];
     
            // Update the number used
            v3[A[i] - 1] = v2[i + 1];
        }
    }
     
    // Last answer is stored in v2
    return v2[0];
}
 
// Driver Code
let a = [ 1, 2, 3, 4 ];
let n = a.length;
let k = 3;
 
document.write(solution(a, n, k));
 
// This code is contributed by avanitrachhadiya2155
 
</script>


Output

4

Complexity Analysis:

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


Like Article
Suggest improvement
Previous
Next
Share your thoughts in the comments

Similar Reads