397. Integer Replacement

MediumGreedyBit ManipulationMemoizationDynamic Programming

Problem Description

Given a positive integer n, the task is to transform this number to 1. There are two types of operations that you are allowed to perform:

If n is even, you can divide it by 2 (n = n / 2).
If n is odd, you can either increase it by 1 (n = n + 1) or decrease it by 1 (n = n - 1).

Each operation counts as a single step, and the goal is to find the minimum number of steps required to reduce n to 1.

Intuition

The intuition behind the solution involves understanding parity (odd or even nature) of n and using bitwise operations for efficiency. For an even n, the decision is straightforward: divide by 2. This halving operation is efficient because it significantly reduces n.

For an odd n, the decision to subtract or add 1 relies on considering the bits of n. Specifically, we observe the following patterns:

When the least significant two bits of n are 01 (meaning the next division will yield an even number), it is often better to subtract 1 than to add 1, because we want to get to an even number to divide by 2 in the next step.
However, there is a special case when n is 3; the best operation is to subtract 1 twice to reach 1 rather than add 1 and then divide twice.
If the least significant two bits are 11, adding 1 converts n to an even number, which can then be halved. This is except for the case when n is 3, as mentioned above.

The approach makes use of bitwise operations:

We use n & 1 to check if n is even or odd (0 means even and 1 means odd).
The right shift n >>= 1 effectively divides n by 2.
We check if the last two bits are 11 by n & 3, and if the result is 3, we know adding 1 is a favorable operation.

The solution employs a loop that iteratively applies the optimal operation until n becomes 1, and a counter ans keeps track of the number of operations performed.

Learn more about Greedy, Memoization and Dynamic Programming patterns.

Not Sure What to Study? Take the 2-min Quiz to Find Your Missing Piece：

Solution Approach

The solution provided uses a greedy approach to minimize the number of operations needed to reduce n to 1 with the two allowed operations, substituting and dividing.

Here's a walkthrough of the algorithm using the Reference Solution Approach:

Initialize ans to 0. This variable will keep track of the total number of operations performed.
Enter a loop that continues until n is equal to 1.
Inside the loop, check if n is even by using the bitwise AND operation n & 1. If the result is 0, it means n is even.
If n is even, we apply the bitwise right shift operation n >>= 1 to n. This operation is equivalent to dividing n by 2.
If n is odd, we have to decide whether to increment or decrement n. To make this decision:
- Check if n is not equal to 3 and if the last two bits of n are 11 by applying the bitwise AND operation n & 3. If the result is 3, it's better to increment n by 1, because doing so will lead to an even number after this operation.
- In all other cases (when n is odd, and either n is 3 or the last two bits of n are not 11), it's better to decrement n by 1.
After each operation (incrementing, decrementing, or dividing), increment ans by 1 to count the operation.
Once n is reduced to 1, exit the loop, and return ans as the total number of operations performed.

The solution applies mathematical operations and understands binary representations to guide the decision-making process. No additional data structures are used, and the algorithm runs in a time complexity that is logarithmic to the value of n, specifically O(log n), because each division by two halves the problem size.

Discover Your Strengths and Weaknesses: Take Our 2-Minute Quiz to Tailor Your Study Plan:

Example Walkthrough

Let's illustrate the solution approach with a small example using n = 15.

Initialize ans to 0. At the start, no steps have been taken yet, so ans is 0.
Because n is not 1, we enter the loop.
As n is odd (15 & 1 is 1), we proceed to check if we should decrement or increment n.
Since n equals 15, which is not 3, and the last two bits of n are 11 (15 & 3 gives 3), we increment n by 1, resulting in n = 16. Now ans = 1.
Now n is even (16 & 1 is 0), so we right shift n, effectively dividing it by 2. n becomes 8 (n >>= 1). Now ans = 2.
n is still even, so we keep right shifting. n becomes 4. Now ans = 3.
Continue with the even case; n becomes 2. Now ans = 4.
Finally, n is 2, which is even again, and after one more shift, n becomes 1. Now ans = 5.
Since n is now equal to 1, we break out of the loop.
The value of ans is 5, representing the minimum number of steps needed to transform n from 15 to 1.

The solution successfully applies the steps from the Reference Solution Approach, selecting operations that gradually reduce n to 1 with optimal efficiency, ending with an answer of 5 steps for this example.

Solution Implementation

1class Solution:
2    def integer_replacement(self, n: int) -> int:
3        # Initialize a counter for the number of steps taken
4        step_count = 0
5      
6        # Continue processing until the integer becomes 1
7        while n != 1:
8            # If n is even, shift it right by 1 (equivalent to dividing by 2)
9            if (n & 1) == 0:
10                n >>= 1
11            # If n is one less than a multiple of 4 (except when n is 3)
12            # increment n (e.g., for 7 -> 8 is better than 7 -> 6)
13            elif n != 3 and (n & 3) == 3:
14                n += 1
15            # In all other cases (n is odd and not handled by the above condition), decrement n
16            else:
17                n -= 1
18            # Increment the step count after each operation
19            step_count += 1
20      
21        # Return the total number of steps taken
22        return step_count
23

1class Solution {
2    public int integerReplacement(int n) {
3        int steps = 0; // Counter for the number of steps taken to transform 'n' to 1
4      
5        while (n != 1) {
6            if ((n & 1) == 0) { // If 'n' is even
7                n >>>= 1; // Right shift (unsigned) to divide 'n' by 2
8            } else if (n != 3 && (n & 3) == 3) { // If 'n' is not 3 and the last two bits are 11
9                n++; // Increment 'n' since it leads to more 0s when it's divided by 2 subsequently 
10            } else {
11                n--; // Decrement 'n' if it's odd and doesn't match the previous case
12            }
13            steps++; // Increment the step count after each operation
14        }
15        return steps; // Return the total number of steps once 'n' is reduced to 1
16    }
17}
18

1class Solution {
2public:
3    // Function to determine the minimum number of operations to transform
4    // a given integer n to 1 by either decrementing by 1, incrementing by 1,
5    // or halving it when even.
6    int integerReplacement(int n) {
7        int operationsCount = 0; // variable to count the number of operations
8        long longNumber = n; // use a long to handle overflow
9      
10        // Continue until longNumber becomes 1
11        while (longNumber != 1) {
12            if ((longNumber & 1) == 0) {
13                // If longNumber is even, halve it
14                longNumber >>= 1; 
15            } else if (longNumber != 3 && (longNumber & 3) == 3) {
16                // If longNumber is not 3 and ends with binary '11', increment it
17                longNumber++;
18            } else {
19                // If longNumber is odd and not covered by the above case, decrement it
20                longNumber--;
21            }
22            // Increment the operation count after each operation
23            operationsCount++;
24        }
25      
26        // Return the total count of operations performed
27        return operationsCount;
28    }
29};
30

1// Global variable to count the number of operations
2let operationsCount: number = 0;
3
4// Function to determine the minimum number of operations to transform
5// a given integer n to 1 by either decrementing by 1, incrementing by 1,
6// or halving it when even.
7function integerReplacement(n: number): number {
8    // Initialize count and use a 'BigInt' for 'n' to handle overflow
9    operationsCount = 0;
10    let longNumber: bigint = BigInt(n);
11  
12    // Continue processing until longNumber becomes 1
13    while (longNumber !== BigInt(1)) {
14        if ((longNumber & BigInt(1)) === BigInt(0)) {
15            // If longNumber is even, right shift equals dividing by 2
16            longNumber >>= BigInt(1);
17        } else if (longNumber !== BigInt(3) && (longNumber & BigInt(3)) === BigInt(3)) {
18            // If longNumber ends with binary '11' (is odd) and is not 3, increment
19            longNumber++;
20        } else {
21            // If longNumber is odd and not covered by the above case, decrement
22            longNumber--;
23        }
24        // Increment the operation count after each operation
25        operationsCount++;
26    }
27  
28    // Return the total count of operations performed
29    return operationsCount;
30}
31
32// Example usage:
33// To call the function, simply invoke it with an integer argument
34// let result: number = integerReplacement(1234);
35

Not Sure What to Study? Take the 2-min Quiz：

Time and Space Complexity

The given code implements an algorithm to determine the minimum number of operations to reduce a number n to 1 by either dividing it by 2 if it's even or subtracting 1 or adding 1 if it's odd, which converges to division by 2 when possible.

Time Complexity:

The time complexity of the algorithm is determined by the number of operations needed to reduce n to 1. In the worst case scenario, for each bit in the binary representation of n, the loop might be executed twice (once for subtraction/addition to make it even, and once for the division by 2). However, the addition operation in n += 1 can lead to the removal of multiple trailing '1's in binary, which means it's possible to skip several steps at once. Therefore, the time complexity is O(log n) in the average case, and in the worst case, it's slightly more than O(log n) due to the possible addition steps that can reduce the number of 1's in the binary representation.

Space Complexity:

The space complexity of the code is O(1), as the algorithm uses a fixed amount of space - the variable ans to keep count of the operations, and n is modified in place without using additional memory that scales with the input size.

Learn more about how to find time and space complexity quickly using problem constraints.

Fast Track Your Learning with Our Quick Skills Quiz:

What is the best way of checking if an element exists in a sorted array once in terms of time complexity? Select the best that applies.

Linear SearchBinary SearchQuick SelectHash Set

I don’t know

Overview

Basic Data Structures

Basic Algorithms

Overview

Sorted Array

Implicitly Sorted Array

Advanced

Speedrun

Introduction

DFS on Tree

Binary Search Tree

Advanced

Combinatorial Search

Pruning

Additional States

Aggregation and Memoization

Dedup

Additional Practices

Speedrun

Introduction

BFS on Tree

Introduction

Vanilla BFS

Matrix as Graph

Implicit Graph

Directed Graph / Topological Sort

Weighted Graph

Minimum Spanning Tree

Speedrun

Introduction

Same Direction

Opposite Direction

Sliding Window

Prefix Sum

Cycle Finding

Advanced

Speedrun

Introduction

Top K

Moving Best

Multiple Heaps

Introduction

Grid

Dynamic number of subproblems

Interval

Dual-Sequence

Knapsack

Bitmask

Additional Practices

Disjoint Set Union | Union Find

Trie

Data Structure Design

Segment Tree

Interval

Monotonic Stack

Divide and Conquer

Line Sweep

Tree Traversal without Recursion

Greedy

Math

Matrix

Classics

Concepts

Projects

Amazon OA

Microsoft OA

Google OA

Twitter OA

Editorials

397. Integer Replacement

Problem Description

Intuition

Solution Approach

Example Walkthrough

Solution Implementation

Time and Space Complexity

Time Complexity:

Space Complexity:

Recommended Readings