Hard
题目描述
给你两个长度相同的正整数数组 nums 和 target。
在一次操作中,你可以选择 nums 的任意子数组,并将该子数组中的每个元素加 1 或减 1。
返回使 nums 等于数组 target 所需的最少操作数。
示例 1:
输入:nums = [3,5,1,2], target = [4,6,2,4]
输出:2
解释:
我们将执行以下操作使 nums 等于 target:
- 将 nums[0..3] 增加 1,nums = [4,6,2,3]。
- 将 nums[3..3] 增加 1,nums = [4,6,2,4]。
示例 2:
输入:nums = [1,3,2], target = [2,1,4]
输出:5
解释:
我们将执行以下操作使 nums 等于 target:
- 将 nums[0..0] 增加 1,nums = [2,3,2]。
- 将 nums[1..1] 减少 1,nums = [2,2,2]。
- 将 nums[1..1] 减少 1,nums = [2,1,2]。
- 将 nums[2..2] 增加 1,nums = [2,1,3]。
- 将 nums[2..2] 增加 1,nums = [2,1,4]。
提示:
1 <= nums.length == target.length <= 10^51 <= nums[i], target[i] <= 10^8
解题思路
解题思路
这道题的关键在于理解操作的本质:我们可以选择任意子数组进行统一的加1或减1操作。
核心观察:
- 将问题转化:设
diff[i] = nums[i] - target[i],目标是让所有diff[i]都变为0 - 每次操作相当于选择一个子数组,将其中所有元素统一加1或减1
贪心策略: 使用分治思想。对于数组中的每个位置,我们考虑以下几种情况:
- 如果当前位置的
diff值为0,则不需要操作 - 如果当前位置的
diff值与前一位置同号,可以复用之前的操作 - 如果当前位置的
diff值与前一位置异号,需要新增操作
算法流程:
- 计算差值数组
diff[i] = nums[i] - target[i] - 遍历差值数组,对于每个位置:
- 如果与前一位置同号,可以复用操作,只需要处理差值的增量部分
- 如果异号或符号改变,需要独立处理
- 累加所有必需的操作次数
时间复杂度:O(n),空间复杂度:O(1)
代码实现
class Solution {
public:
long long minimumOperations(vector<int>& nums, vector<int>& target) {
int n = nums.size();
long long operations = 0;
long long prev_diff = 0;
for (int i = 0; i < n; i++) {
long long curr_diff = (long long)nums[i] - target[i];
if (i == 0) {
operations += abs(curr_diff);
} else {
// 如果当前差值与前一个差值同号,可以复用操作
if ((curr_diff > 0 && prev_diff > 0) || (curr_diff < 0 && prev_diff < 0)) {
// 只需要处理增量部分
if (abs(curr_diff) > abs(prev_diff)) {
operations += abs(curr_diff) - abs(prev_diff);
}
} else {
// 异号或者从0开始,需要独立操作
operations += abs(curr_diff);
}
}
prev_diff = curr_diff;
}
return operations;
}
};
class Solution:
def minimumOperations(self, nums: List[int], target: List[int]) -> int:
n = len(nums)
operations = 0
prev_diff = 0
for i in range(n):
curr_diff = nums[i] - target[i]
if i == 0:
operations += abs(curr_diff)
else:
# 如果当前差值与前一个差值同号,可以复用操作
if (curr_diff > 0 and prev_diff > 0) or (curr_diff < 0 and prev_diff < 0):
# 只需要处理增量部分
if abs(curr_diff) > abs(prev_diff):
operations += abs(curr_diff) - abs(prev_diff)
else:
# 异号或者从0开始,需要独立操作
operations += abs(curr_diff)
prev_diff = curr_diff
return operations
public class Solution {
public long MinimumOperations(int[] nums, int[] target) {
int n = nums.Length;
long operations = 0;
long prevDiff = 0;
for (int i = 0; i < n; i++) {
long currDiff = (long)nums[i] - target[i];
if (i == 0) {
operations += Math.Abs(currDiff);
} else {
// 如果当前差值与前一个差值同号,可以复用操作
if ((currDiff > 0 && prevDiff > 0) || (currDiff < 0 && prevDiff < 0)) {
// 只需要处理增量部分
if (Math.Abs(currDiff) > Math.Abs(prevDiff)) {
operations += Math.Abs(currDiff) - Math.Abs(prevDiff);
}
} else {
// 异号或者从0开始,需要独立操作
operations += Math.Abs(currDiff);
}
}
prevDiff = currDiff;
}
return operations;
}
}
/**
* @param {number[]} nums
* @param {number[]} target
* @return {number}
*/
var minimumOperations = function(nums, target) {
const n = nums.length;
const diff = new Array(n);
for (let i = 0; i < n; i++) {
diff[i] = target[i] - nums[i];
}
let operations = 0;
let stack = 0;
for (let i = 0; i < n; i++) {
if (diff[i] > stack) {
operations += diff[i] - stack;
} else if (diff[i] < stack) {
operations += stack - diff[i];
}
stack = diff[i];
}
operations += Math.abs(stack);
return operations;
};
复杂度分析
| 复杂度类型 | 复杂度 |
|---|---|
| 时间复杂度 | O(n) |
| 空间复杂度 | O(1) |