题目
给定一个数组,包含从 1 到 N 所有的整数,但其中缺了两个数字。你能在 O(N) 时间内只用 O(1) 的空间找到它们吗?
以任意顺序返回这两个数字均可。
示例 1:
输入: [1]
输出: [2,3]
示例 2:
输入: [2,3]
输出: [1,4]
提示:
nums.length <= 30000
题解1
思路
- 将数组排序,若 n u m s [ i ] − n u m s [ i − 1 ] ≠ 1 nums[i]-nums[i-1] \ne 1 nums[i]−nums[i−1]=1则说明 n u m s [ i ] − 1 nums[i]-1 nums[i]−1不在数组中
- 同时将消失的数字在数组首和数组尾两种情况单独考虑
代码
class Solution:
def missingTwo(self, nums: List[int]) -> List[int]:
if not nums: return [1,2]
elif len(nums) == 1:
if nums[0] == 1: return [2,3]
elif nums[0] == 2: return [1,3]
elif nums[0] == 3: return [1,2]
else: assert False
nums.sort()
ret = []
if nums[0] > 2: return [1,2]
elif nums[0] == 2: ret.append(1)
i = 1
while i < len(nums):
if nums[i] - nums[i-1] > 1:
ret.append(nums[i] - 1)
i+=1
if len(ret) == 0:
ret.append(nums[-1] + 1)
ret.append(nums[-1] + 2)
elif len(ret) == 1:
ret.append(nums[-1] + 1)
return ret
复杂度
- 时间复杂度: O ( n log n ) O(n\log n) O(nlogn)
- 空间复杂度: O ( log n ) O(\log n) O(logn)
题解2
思路
- 数组的长度为 n n n则说明 [ 1 , n + 2 ] [1,n+2] [1,n+2]区间内,不曾出现过的数则说明消失了
代码
class Solution:
def missingTwo(self, nums: List[int]) -> List[int]:
visited = [False] * (len(nums)+2)
for n in nums:
visited[n-1] = True
ret = []
for i in range(len(visited)):
if not visited[i]: ret.append(i+1)
return ret
复杂度
- 时间复杂度: O ( n ) O(n) O(n)
- 空间复杂度: O ( n ) O(n) O(n)
题解3
思路
- 假设数组消失的数为 x 1 , x 2 x_1, x_2 x1,x2,则 x 1 + x 2 = ∑ n u m − ∑ ( r a n g e ( 1 , n + 2 ) ) x_1+x_2=\sum num - \sum(range(1, n+2)) x1+x2=∑num−∑(range(1,n+2))
- 若输入为哈希表,则可以通过 x 1 ∉ n u m s ∧ x 2 ∉ n u m s x_1 \notin nums \land x_2\notin nums x1∈/nums∧x2∈/nums来判断 可惜不是╮(╯▽╰)╭
代码
class Solution:
def missingTwo(self, nums: List[int]) -> List[int]:
s = set(nums)
n = len(s)
currentSum = sum(s)
expectSum = (n+2)*(n+3)//2
diff = expectSum - currentSum
for i in range(1, n+3):
if i not in s and (diff - i) not in s and (diff-i)<n + 3: return [i, diff-i]
复杂度
- 时间复杂度: O ( n ) O(n) O(n)
- 空间复杂度: O ( n ) O(n) O(n)
题解4
思路
-
继题解3,考虑构造线性无关方程组,考虑:
{ a = x 1 + x 2 = ( n + 2 ) ( n + 3 ) 2 − ∑ ( n u m s ) b = x 1 2 + x 2 2 = ( n + 2 ) ( n + 3 ) ( 2 n + 5 ) 6 − ∑ n ∈ n u m s ( n 2 ) \left\{\begin{aligned} & a = x_1+x_2 = \frac{(n+2)(n+3)}{2} - \sum(nums)\\ & b = x_1^2 + x_2^2 = \frac{(n+2)(n+3)(2n+5)}{6} -\sum_{n \ \in nums}(n^2)\\ \end{aligned}\right. ⎩ ⎨ ⎧a=x1+x2=2(n+2)(n+3)−∑(nums)b=x12+x22=6(n+2)(n+3)(2n+5)−n ∈nums∑(n2) -
可以求解
δ = 2 b − a 2 { x 1 = a + δ 2 x 1 = a − δ 2 \delta = \sqrt{2b-a^2}\\ \left\{\begin{aligned} & x_1 = \frac{a + \delta}{2} \\ & x_1 = \frac{a - \delta}{2} \end{aligned}\right. δ=2b−a2⎩ ⎨ ⎧x1=2a+δx1=2a−δ
代码
class Solution:
def missingTwo(self, nums: List[int]) -> List[int]:
n = len(nums)
currentSum = sum(nums)
currentSquareSum = sum(map(lambda x: x*x, nums))
expectSum = (n+2)*(n+3)//2
expectSquareSum = (n+2)*(n+3)*(2*n+5)//6
a = expectSum - currentSum
b = expectSquareSum - currentSquareSum
delta = math.sqrt(2 * b - a * a)
return [int((a + delta)//2), int((a-delta)//2)]
- 时间复杂度: O ( n ) O(n) O(n)
- 空间复杂度: O ( 1 ) O(1) O(1)
题解5
思路
- 对所有数和 [ 1 , n + 2 ] [1,n+2] [1,n+2]中的所有数位运算,最后得到的结果为 x 1 ⊕ x 2 x_1 \oplus x_2 x1⊕x2
- 因为python中负数使用补码表示,所以可以使用 x o r s u m ⊕ − x o r s u m xorsum \oplus -xorsum xorsum⊕−xorsum作为掩码进行分类
- 分组后再通过掩码计算分别求出两个不同的数
代码
class Solution:
def missingTwo(self, nums: List[int]) -> List[int]:
xorSum = 0
for n in nums:
xorSum ^= n
for n in range(1, len(nums) + 3):
xorSum ^= n
mask = xorSum & (-xorSum)
n1, n2 = 0,0
for n in nums:
if n&mask:
n1 ^= n
else:
n2 ^= n
for n in range(1, len(nums) + 3):
if n&mask:
n1 ^= n
else:
n2 ^= n
return [n1,n2]
复杂度
- 时间复杂度: O ( n ) O(n) O(n)
- 空间复杂度: O ( 1 ) O(1) O(1)