Count of Binary strings of length N having atmost M consecutive 1s or 0s alternatively exactly K times

Difficulty Level : Medium
Last Updated : 03 May, 2021

Given three integers, N, K and M. The task is to find out the number of binary strings of length N which always starts with 1, in which there can be at most M consecutive 1’s or 0’s and they alternate exactly K times.
Examples:

Input: N = 5, K = 3, M = 2
Output: 3
The 3 configurations are:
11001
10011
11011
Explanation:
Notice that the groups of 1’s and 0’s alternate exactly K times
Input: N = 7, K = 4, M = 3
Output: 16

Approach: Since this problem involves both overlapping sub-problem and optimal substructure. So, this problem can be solved using dynamic programming.

Sub-problem: DP[i][j] represents the number of binary strings upto length i having j alternating groups till now. So, to calculate dp[N][K] if we know the value of dp[n-j][k-1], then we can easily get the result by summing up the sub-problem value over j = 1 to m (DP[N][K] represents the final answer).
As shown below in the recursion tree diagram, it is observed many sub-problem overlaps. So, the result needs to be cached to avoid redundant calculations.

Optimal substructure:
$$dp[i][j]=$$\sum_{j=1}^{M} f(N-j, K-1)$
By following the top-down DP approach:
As we can have a group which can be atmost of the length M, so we iterate on every possible length and recur with new N and decreasing K by 1, as a new group is formed. Solution to sub-problem is cached and summed up to give final result dp[N][K].
Base Case:
1. When N is 0 and K is 0, then return 1
2. When N is 0 but K is not 0, then return 0
3. When N is not 0 but K is 0, then return 0
4. When both are negative, return 0

Below is the implementation of above approach:

C++

// C++ program to find the count
// of Binary strings of length N
// having atmost M consecutive 1s or 0s
// alternatively exactly K times
 
#include <bits/stdc++.h>
using namespace std;
 
// Array to contain the final result
int dp[1000][1000];
 
// Function to get the number
// of desirable binary strings
int solve(int n, int k, int m)
{
 
    // if we reach end of string
    // and groups are exhausted,
    // return 1
    if (n == 0 && k == 0)
        return 1;
 
    // if length is exhausted but
    // groups are still to be made,
    // return 0
    if (n == 0 && k != 0)
        return 0;
 
    // if length is not exhausted
    // but groups are exhausted,
    // return 0
    if (n != 0 && k == 0)
        return 0;
 
    // if both are negative
    // just return 0
    if (n < 0 || k < 0)
        return 0;
 
    // if already calculated,
    // return it
    if (dp[n][k])
        return dp[n][k];
 
    // initialise answer
    // for each state
    int ans = 0;
 
    // loop through every
    // possible m
    for (int j = 1; j <= m; j++) {
        ans += solve(n - j, k - 1, m);
    }
    return dp[n][k] = ans;
}
 
// Driver code
int main()
{
 
    int N = 7, K = 4, M = 3;
    cout << solve(N, K, M);
}

Java

// Java program to find the count of
// Binary Strings of length N having
// atmost M consecutive 1s or 0s
// alternatively exactly K times
import java.util.*;
 
class GFG{
 
// Array to contain the final result
static int [][]dp = new int[1000][1000];
 
// Function to get the number
// of desirable binary strings
static int solve(int n, int k, int m)
{
 
    // If we reach end of string
    // and groups are exhausted,
    // return 1
    if (n == 0 && k == 0)
        return 1;
 
    // If length is exhausted but
    // groups are still to be made,
    // return 0
    if (n == 0 && k != 0)
        return 0;
 
    // If length is not exhausted
    // but groups are exhausted,
    // return 0
    if (n != 0 && k == 0)
        return 0;
 
    // If both are negative
    // just return 0
    if (n < 0 || k < 0)
        return 0;
 
    // If already calculated,
    // return it
    if (dp[n][k] > 0)
        return dp[n][k];
 
    // Initialise answer
    // for each state
    int ans = 0;
 
    // Loop through every
    // possible m
    for(int j = 1; j <= m; j++)
    {
       ans += solve(n - j, k - 1, m);
    }
    return dp[n][k] = ans;
}
 
// Driver code
public static void main(String[] args)
{
    int N = 7, K = 4, M = 3;
    System.out.print(solve(N, K, M));
}
}
 
// This code is contributed by Rajput-Ji

Python 3

# Python3 program to find the count
# of Binary strings of length N
# having atmost M consecutive 1s or
# 0s alternatively exactly K times
 
# List to contain the final result
rows, cols = (1000, 1000)
dp = [[0 for i in range(cols)]
         for j in range(rows)]
 
# Function to get the number
# of desirable binary strings
def solve(n, k, m):
     
    # If we reach end of string
    # and groups are exhausted,
    # return 1
    if n == 0 and k == 0:
        return 1
 
    # If length is exhausted but
    # groups are still to be made,
    # return 0
    if n == 0 and k != 0:
        return 0
 
    # If length is not exhausted
    # but groups are exhausted,
    # return 0
    if n != 0 and k == 0:
        return 0
 
    # If both are negative
    # just return 0
    if n < 0 or k < 0:
        return 0
 
    # If already calculated,
    # return it
    if dp[n][k]:
        return dp[n][k]
 
    # Initialise answer
    # for each state
    ans = 0
 
    # Loop through every
    # possible m
    for j in range(1, m + 1):
        ans = ans + solve(n - j,
                          k - 1, m)
    dp[n][k] = ans
     
    return dp[n][k]
 
# Driver code
N = 7
K = 4
M = 3
 
print(solve(N, K, M))
 
# This code is contributed by ishayadav181

C#

// C# program to find the count of
// binary strings of length N having
// atmost M consecutive 1s or 0s
// alternatively exactly K times
using System;
 
class GFG{
 
// Array to contain the readonly result
static int [,]dp = new int[1000, 1000];
 
// Function to get the number
// of desirable binary strings
static int solve(int n, int k, int m)
{
 
    // If we reach end of string
    // and groups are exhausted,
    // return 1
    if (n == 0 && k == 0)
        return 1;
 
    // If length is exhausted but
    // groups are still to be made,
    // return 0
    if (n == 0 && k != 0)
        return 0;
 
    // If length is not exhausted
    // but groups are exhausted,
    // return 0
    if (n != 0 && k == 0)
        return 0;
 
    // If both are negative
    // just return 0
    if (n < 0 || k < 0)
        return 0;
 
    // If already calculated,
    // return it
    if (dp[n, k] > 0)
        return dp[n, k];
 
    // Initialise answer
    // for each state
    int ans = 0;
 
    // Loop through every
    // possible m
    for(int j = 1; j <= m; j++)
    {
       ans += solve(n - j, k - 1, m);
    }
    return dp[n, k] = ans;
}
 
// Driver code
public static void Main(String[] args)
{
    int N = 7, K = 4, M = 3;
     
    Console.Write(solve(N, K, M));
}
}
 
// This code is contributed by gauravrajput1

Javascript

<script>
 
// Javascript program to find the count of
// Binary Strings of length N having
// atmost M consecutive 1s or 0s
// alternatively exactly K times
 
    // Array to contain the final result
     dp = Array(1000);
     for(i =0;i<1000;i++)
         dp[i] = Array(1000).fill(0);
 
    // Function to get the number
    // of desirable binary strings
    function solve(n , k , m)
    {
 
        // If we reach end of string
        // and groups are exhausted,
        // return 1
        if (n == 0 && k == 0)
            return 1;
 
        // If length is exhausted but
        // groups are still to be made,
        // return 0
        if (n == 0 && k != 0)
            return 0;
 
        // If length is not exhausted
        // but groups are exhausted,
        // return 0
        if (n != 0 && k == 0){
            return 0;
            }
 
        // If both are negative
        // just return 0
        if (n < 0 || k < 0)
            return 0;
 
        // If already calculated,
        // return it
        if (dp[n][k] > 0)
            return dp[n][k];
 
        // Initialise answer
        // for each state
        var ans = 0;
 
        // Loop through every
        // possible m
        for (var j = 1; j <= m; j++) {
            ans += solve(n - j, k - 1, m);
           // document.write(ans);
        }
        return dp[n][k] = ans;
    }
 
    // Driver code
     
        var N = 7, K = 4, M = 3;
        document.write(solve(N, K, M));
 
// This code contributed by umadevi9616
 
</script>

Output:

Time complexity: O(N*K*M)

Attention reader! Don’t stop learning now. Get hold of all the important DSA concepts with the DSA Self Paced Course at a student-friendly price and become industry ready. To complete your preparation from learning a language to DS Algo and many more, please refer Complete Interview Preparation Course.

In case you wish to attend live classes with experts, please refer DSA Live Classes for Working Professionals and Competitive Programming Live for Students.

My Personal Notes