2289. Steps to Make Array Non-decreasing

MediumStackArrayLinked ListMonotonic Stack

Problem Description

In this problem, we are given an array nums which is 0-indexed, meaning the first element is at position 0. We perform a certain process repeatedly on this array. In a single step of this process, we remove every element nums[i] if it is smaller than its predecessor nums[i - 1], for all positions i that are greater than 0. We continue performing this step until the array becomes non-decreasing, meaning each element is greater than or equal to the one before it. The task is to calculate the total number of steps required to make the array non-decreasing.

Quick Interview Experience

Help others by sharing your interview experience

Have you seen this problem before?

Intuition

The intuition behind the solution can stem from realizing that elements in the array only need to be considered for removal if they are less than some element that previously appeared. One efficient way to keep track of which elements could possibly be removed is to use a stack. The stack is a data structure allowing us to store elements and remove the tops if conditions are met (in this case, if an element is greater than the elements on top of it).

We iterate the array from right to left because we want to know which elements on the right (upcoming elements in this backward traversal) can be removed. Each step of the iteration is an opportunity to either push a new element onto the stack or to remove one or more elements that are smaller than the current element being considered. To avoid removing elements multiple times, we use a dp (dynamic programming) array to keep track of the number of steps taken to remove the next smaller elements.

When we hit an element that could cause the removal of other elements, we need to calculate how many removal steps it took for the next-smallest elements and update it to the maximum of the current steps recorded plus one or the steps taken for that element already. This ensures we always have the maximum number of steps it would take for each element to be removed, considering all its future smaller elements.

The solution completes when we have considered every element and the dp array contains the steps required to make each individual element non-decreasing in the context of the elements to its right. The answer to the problem is the maximum value in the dp array because that will be the maximum number of steps required to make any element satisfy the non-decreasing property in the context of the entire array.

Learn more about Stack, Linked List and Monotonic Stack patterns.

Solution Approach

The implementation of the solution is based on a stack-based approach coupled with dynamic programming. Here are the key parts of the implementation:

We initialize a stack named stk to keep track of the indices of elements that potentially need to be removed. This stack helps us to maintain the current sequence of elements under consideration for removal in descending order.
Additionally, we create an array dp of the same size as the input array nums. Each entry dp[i] represents the number of steps required to remove the element at nums[i] due to a larger preceding element.
We then iterate through the array nums in reverse (from right to left). At each index i, we consider whether nums[i] can lead to the removal of elements that are currently on the stack.
Inside the loop, we perform the following actions:
- While the stack is not empty and the top element of the stack (the element at the index stk[-1]) is less than nums[i], it means that nums[i] can lead to the removal of nums[stk[-1]]. Therefore, we calculate the dp value for nums[i] which involves incrementing dp[i] by 1 and then comparing it to the dp value of stk[-1], and taking the maximum between them.
- After processing potential removals, i is appended to the stack.
Once we finish iterating through nums, the dp array contains the number of steps each element requires to be removed. The result of the entire operation, which is the number of steps required for nums to become non-decreasing, is the maximum value in the dp array.
In summary, the algorithm effectively processes elements in such a way that it counts how many removal rounds each can survive based on the larger elements before them. This is accomplished by keeping track of the previous elements' survival rounds and updating the current element's count accordingly.
By using a stack, we efficiently manage the sequence of comparisons, only performing removal operations when necessary, and by leveraging dynamic programming, we store intermediate results to avoid redundant calculations.

The concept of dynamic programming, particularly memoization of steps needed for removal, is crucial in avoiding revisiting the same subproblems multiple times, which significantly reduces the overall time complexity of the solution.

Ready to land your dream job?

Unlock your dream job with a 3-minute evaluator for a personalized learning plan!

Start Evaluator

Example Walkthrough

Let's use a small example to illustrate the solution approach. Consider the array nums = [5, 3, 7, 6, 2].

We initialize an empty stack stk and an array dp of the same size as nums initialized to 0s, representing the number of steps needed to remove each element. So in this case, dp = [0, 0, 0, 0, 0].
We start iterating through the array nums from right to left:

i. For i = 4 (nums[4] = 2), there are no previous elements, so we push i onto stk which now becomes [4].

ii. Move to i = 3 (nums[3] = 6), stk is not empty and nums[stk[-1]] = 2 < 6, so we pop 4 from stk, update dp[3] = max(dp[3]+1, dp[4]) ⇒ dp[3] = 1. stk becomes empty and we push 3 onto stk - stack is [3].

iii. Move to i = 2 (nums[2] = 7), stk is not empty and nums[stk[-1]] = 6 < 7, so we pop 3 from stk, update dp[2] = max(dp[2]+1, dp[3]) ⇒ dp[2] = 2. stk becomes empty and we push 2 onto stk - stack is [2].

iv. Move to i = 1 (nums[1] = 3), stk is not empty but nums[stk[-1]] = 7 > 3, no need to pop 2 from stk, just push 1 onto stk - stack is now [2, 1].

v. Move to i = 0 (nums[0] = 5), stk is not empty and nums[stk[-1]] = 3 < 5, so we pop 1 from the stack, update dp[0] = max(dp[0]+1, dp[1]) ⇒ dp[0] = 1. Since nums[stk[-1]] = 7 > 5, we stop popping from the stack and push 0 onto the stack - stk becomes [2, 0].
After iterating, our dp array has the values [1, 0, 2, 1, 0]. Each entry in dp represents the number of rounds an element can survive (0 indicates it doesn't need to be removed).
We look for the maximum value in the dp array, which is dp[2] = 2, meaning the total number of steps required to make the array non-decreasing is 2. These steps translate to:
- Step 1: Remove nums[3] (6) because it is less than nums[2] (7). Array becomes [5, 3, 7, 2].
- Step 2: Remove nums[1] (3) and nums[3] (2) because they are less than nums[0] (5) and nums[2] (7), respectively. Array becomes [5, 7].

The result indicates that the original array [5, 3, 7, 6, 2] can be made non-decreasing in 2 steps, resulting in the array [5, 7].

Solution Implementation

1class Solution:
2    def totalSteps(self, nums: List[int]) -> int:
3        # Initialize an empty stack to keep track of indices
4        stack = []
5      
6        # Use this list to keep track of the number of steps
7        # required to remove each number
8        steps_to_remove = [0] * len(nums)
9      
10        # Initialize the max_steps as 0, to store the maximum
11        # number of steps needed to remove any element
12        max_steps = 0
13
14        # Iterate the nums list in reverse order
15        for i in range(len(nums) - 1, -1, -1):
16            # While the stack is not empty and the current number is greater than 
17            # the top element of the stack
18            while stack and nums[i] > nums[stack[-1]]:
19                # Calculate the steps to remove the current number
20                # It's either 1 step more than its last or the steps needed to remove 
21                # the number at the top of the stack, whichever is greater
22                steps_to_remove[i] = max(steps_to_remove[i] + 1, steps_to_remove[stack.pop()])
23          
24            # Add the current index to the stack
25            stack.append(i)
26      
27        # Return the maximum steps needed to remove an element from nums
28        max_steps = max(steps_to_remove)
29        return max_steps
30

1class Solution {
2  
3    /**
4     * Calculate the total steps needed to remove all possible elements according to the game rules.
5     * 
6     * @param nums The array of integers representing the initial sequence.
7     * @return The total number of steps to remove all possible elements.
8     */
9    public int totalSteps(int[] nums) {
10        // Creating a stack to keep track of indices of elements in 'nums'
11        Deque<Integer> stack = new ArrayDeque<>();
12      
13        // Initialize the answer (total steps) to 0
14        int totalSteps = 0;
15      
16        // The length of the input array
17        int length = nums.length;
18      
19        // Creating an array to store the number of rounds each element can survive before being removed
20        int[] rounds = new int[length];
21      
22        // Iterate over the elements in reverse
23        for (int i = length - 1; i >= 0; --i) {
24            // While stack is not empty and the current element is greater than the next element in the stack
25            while (!stack.isEmpty() && nums[i] > nums[stack.peek()]) {
26                // Calculate the number of rounds needed for the current element
27                // by comparing and taking the maximum between the rounds for the current element 
28                // and the rounds for the element that will be popped from the stack plus one.
29                rounds[i] = Math.max(rounds[i] + 1, rounds[stack.pop()]);
30              
31                // Update the total steps with the maximum number of rounds we've seen
32                totalSteps = Math.max(totalSteps, rounds[i]);
33            }
34          
35            // Push the current index onto the stack
36            stack.push(i);
37        }
38      
39        // Return the total number of rounds (total steps)
40        return totalSteps;
41    }
42}
43

1class Solution {
2public:
3    int totalSteps(vector<int>& nums) {
4        stack<int> indicesStack; // Stack to maintain the indices of nums
5        int maxSteps = 0; // Variable to store the maximum number of steps needed
6        int numsSize = nums.size(); // Size of the nums vector
7        vector<int> steps(numsSize); // Vector to store the number of steps for each element
8      
9        // Iterate over the elements from the end to the beginning
10        for (int i = numsSize - 1; i >= 0; --i) {
11            // Pop elements from the stack while the current element is greater
12            // than the element at the top of the stack
13            while (!indicesStack.empty() && nums[i] > nums[indicesStack.top()]) {
14                // Calculate the steps for the current element based on the existing steps
15                // for the top element and itself
16                steps[i] = max(steps[i] + 1, steps[indicesStack.top()]);
17                // Update maxSteps with the maximum steps needed so far
18                maxSteps = max(maxSteps, steps[i]);
19                // Remove the top element from the stack
20                indicesStack.pop();
21            }
22            // Push the current index onto the stack
23            indicesStack.push(i);
24        }
25        // Return the maximum number of steps required
26        return maxSteps;
27    }
28};
29

1/**
2 * Calculates the maximum number of steps to make the array non-increasing by
3 * removing the minimum element from each non-increasing subarray.
4 * @param nums The array of numbers to be processed.
5 * @return The total number of steps required.
6 */
7function totalSteps(nums: number[]): number {
8    // Initialize the answer to zero.
9    let answer = 0;
10  
11    // Use a stack to keep track of elements and their steps to be removed.
12    let stack: [number, number][] = [];
13  
14    // Iterate over the numbers in the provided array.
15    for (let num of nums) {
16        // Track the maximum number of steps needed so far.
17        let maxSteps = 0;
18      
19        // Remove elements from the stack while the top is less than or equal to the current number.
20        while (stack.length && stack[0][0] <= num) {
21            // Update the maximum steps using the steps value from the element being removed.
22            maxSteps = Math.max(stack[0][1], maxSteps);
23            stack.shift(); // Remove the element from the stack's front.
24        }
25      
26        // If the stack is not empty, increment the maximum steps.
27        if (stack.length) maxSteps++;
28      
29        // Update the global answer with the maximum steps if necessary.
30        answer = Math.max(maxSteps, answer);
31      
32        // Push the current number and its steps onto the stack.
33        stack.unshift([num, maxSteps]);
34    }
35  
36    // Return the calculated answer.
37    return answer;
38}
39

Time and Space Complexity

The given code implements an algorithm that computes the total number of removal steps required for an array of numbers, given certain conditions for removal. The algorithm utilizes a stack and a dynamic programming array to keep track of the maximum steps needed.

Time Complexity:

The time complexity of the algorithm is O(n). Here's why:

We iterate through the list nums in reverse order, which is a linear operation with complexity O(n).
For each element, we perform a while loop that pops elements from the stack until a condition is met. Although it seems like it could lead to a higher complexity, each element is pushed to and popped from the stack at most once. Therefore, the total number of operations for the while loop across all iterations is at most O(n).
The max() function inside the while loop operates in constant time since it's just comparing two integers.

Combining these factors, the complexity remains linear with respect to the length of the input array.

Space Complexity:

The space complexity of the algorithm is O(n) due to the following:

We use a dynamic programming array dp of the same length as the input array nums, which takes O(n) space.
A stack stk is used, which in the worst case could store all elements if they are in increasing order. This also takes O(n) space.

Hence, the overall space complexity is O(n), dominated by the dp array and the stack stk.

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

Discover Your Strengths and Weaknesses: Take Our 3-Minute Quiz to Tailor Your Study Plan:

Which type of traversal does breadth first search do?

Pre-order traversal

Post-order traversal

In-order traversal

Level-order traversal

I don't know