LeetCode 题解工作台

找出不同元素数目差数组

给你一个下标从 0 开始的数组 nums ,数组长度为 n 。 nums 的 不同元素数目差 数组可以用一个长度为 n 的数组 diff 表示,其中 diff[i] 等于前缀 nums[0, ..., i] 中不同元素的数目 减去 后缀 nums[i + 1, ..., n - 1] 中不同元素的数…

category

2

题型

code_blocks

6

代码语言

hub

3

相关题

当前训练重点

简单 · 数组·哈希·扫描

bolt

答案摘要

我们可以预处理出一个后缀数组 ,其中 表示后缀 $nums[i, ..., n - 1]$ 中不同元素的数目,在预处理过程中,我们使用一个哈希表 来维护后缀中出现过的元素,这样我们就可以在 的时间内查询后缀中不同元素的数目。 预处理完后缀数组 后,我们清空哈希表 ,然后再次遍历数组 ,用哈希表 来维护前缀中出现过的元素,那么位置 的答案就是 中不同元素的数目减去 $suf[i + 1…

Interview AiBox logo

Interview AiBox 实时 AI 助手,陪你讲清 数组·哈希·扫描 题型思路

试试 AiBox 面试助手arrow_forward
description

题目描述

给你一个下标从 0 开始的数组 nums ,数组长度为 n

nums不同元素数目差 数组可以用一个长度为 n 的数组 diff 表示,其中 diff[i] 等于前缀 nums[0, ..., i] 中不同元素的数目 减去 后缀 nums[i + 1, ..., n - 1] 中不同元素的数目。

返回 nums不同元素数目差 数组。

注意 nums[i, ..., j] 表示 nums 的一个从下标 i 开始到下标 j 结束的子数组(包含下标 ij 对应元素)。特别需要说明的是,如果 i > j ,则 nums[i, ..., j] 表示一个空子数组。

 

示例 1:

输入:nums = [1,2,3,4,5]
输出:[-3,-1,1,3,5]
解释:
对于 i = 0,前缀中有 1 个不同的元素,而在后缀中有 4 个不同的元素。因此,diff[0] = 1 - 4 = -3 。
对于 i = 1,前缀中有 2 个不同的元素,而在后缀中有 3 个不同的元素。因此,diff[1] = 2 - 3 = -1 。
对于 i = 2,前缀中有 3 个不同的元素,而在后缀中有 2 个不同的元素。因此,diff[2] = 3 - 2 = 1 。
对于 i = 3,前缀中有 4 个不同的元素,而在后缀中有 1 个不同的元素。因此,diff[3] = 4 - 1 = 3 。
对于 i = 4,前缀中有 5 个不同的元素,而在后缀中有 0 个不同的元素。因此,diff[4] = 5 - 0 = 5 。

示例 2:

输入:nums = [3,2,3,4,2]
输出:[-2,-1,0,2,3]
解释:
对于 i = 0,前缀中有 1 个不同的元素,而在后缀中有 3 个不同的元素。因此,diff[0] = 1 - 3 = -2 。
对于 i = 1,前缀中有 2 个不同的元素,而在后缀中有 3 个不同的元素。因此,diff[1] = 2 - 3 = -1 。
对于 i = 2,前缀中有 2 个不同的元素,而在后缀中有 2 个不同的元素。因此,diff[2] = 2 - 2 = 0 。
对于 i = 3,前缀中有 3 个不同的元素,而在后缀中有 1 个不同的元素。因此,diff[3] = 3 - 1 = 2 。
对于 i = 4,前缀中有 3 个不同的元素,而在后缀中有 0 个不同的元素。因此,diff[4] = 3 - 0 = 3 。 

 

提示:

  • 1 <= n == nums.length <= 50
  • 1 <= nums[i] <= 50
lightbulb

解题思路

方法一:哈希表 + 预处理后缀

我们可以预处理出一个后缀数组 sufsuf,其中 suf[i]suf[i] 表示后缀 nums[i,...,n1]nums[i, ..., n - 1] 中不同元素的数目,在预处理过程中,我们使用一个哈希表 ss 来维护后缀中出现过的元素,这样我们就可以在 O(1)O(1) 的时间内查询后缀中不同元素的数目。

预处理完后缀数组 sufsuf 后,我们清空哈希表 ss,然后再次遍历数组 numsnums,用哈希表 ss 来维护前缀中出现过的元素,那么位置 ii 的答案就是 ss 中不同元素的数目减去 suf[i+1]suf[i + 1],即 ssuf[i+1]|s| - suf[i + 1]

时间复杂度 O(n)O(n),空间复杂度 O(n)O(n)。其中 nn 是数组 numsnums 的长度。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
class Solution:
    def distinctDifferenceArray(self, nums: List[int]) -> List[int]:
        n = len(nums)
        suf = [0] * (n + 1)
        s = set()
        for i in range(n - 1, -1, -1):
            s.add(nums[i])
            suf[i] = len(s)
        s.clear()
        ans = [0] * n
        for i, x in enumerate(nums):
            s.add(x)
            ans[i] = len(s) - suf[i + 1]
        return ans
speed

复杂度分析

指标
时间Depends on the final approach
空间Depends on the final approach
psychology

面试官常问的追问

外企场景
  • question_mark

    Focus on how efficiently you can update distinct counts in both prefix and suffix.

  • question_mark

    Watch for candidates who may attempt brute force methods like nested loops that result in O(n^2) time complexity.

  • question_mark

    Evaluate the candidate's approach to optimizing space usage, especially when dealing with hash sets.

warning

常见陷阱

外企场景
  • error

    Using nested loops for counting distinct elements leads to inefficient O(n^2) time complexity.

  • error

    Not updating the suffix count correctly while iterating through the array.

  • error

    Forgetting to handle the case where there are no elements in the suffix for the last element.

swap_horiz

进阶变体

外企场景
  • arrow_right_alt

    Modify the problem to find the distinct difference array for subarrays of different lengths.

  • arrow_right_alt

    Allow negative numbers in the array and handle distinct counting for negative values.

  • arrow_right_alt

    Increase the array size limit and evaluate performance under larger inputs.

help

常见问题

外企场景

找出不同元素数目差数组题解:数组·哈希·扫描 | LeetCode #2670 简单