构造单调队列sortque,使得对当前窗口,sortque中的元素单调递减,首项为窗口最大值;当窗口向右滑动,检查sortque左侧元素是否需要弹出,推入右侧元素时保证新元素左侧的所有元素大于它。空间:维护sortque所用的O(k);时间:遍历一次nums数组,每个元素最多被push一次和pop一次,因此是O(n)。
易错点:j的边界条件判断 & 勿漏最后一个窗口的最大值。
from collections import deque
class sortque:
def __init__(self):
self.que = deque()
def push(self, val):
while self.que and self.que[-1] < val:
self.que.pop()
self.que.append(val)
def pop(self, val):
if self.que[0] == val:
self.que.popleft()
def getMax(self):
if self.que:
return self.que[0]
class Solution:
def maxSlidingWindow(self, nums: List[int], k: int) -> List[int]:
res = []
que = sortque()
for i in range(k):
que.push(nums[i])
for j in range(len(nums) - k):
res.append(que.getMax())
que.pop(nums[j])
que.push(nums[j+k])
res.append(que.getMax())
return res
解法一:map+小顶堆,当大小超过k,移除堆顶,最后剩下的就是前k大的元素。python 中的heapq是小顶堆;若需要大顶堆,可以(-1)*元素再推入堆,返回结果时再*(-1)。
时间O(nlogk),空间O(n)
import heapq
class Solution:
def topKFrequent(self, nums: List[int], k: int) -> List[int]:
#sol1
freq = {}
for num in nums:
freq[num] = freq.get(num, 0) + 1
heap = []
for num, cnt in freq.items():
if len(heap) < k:
heapq.heappush(heap, (cnt, num))
else:
heapq.heappush(heap, (cnt, num))
heapq.heappop(heap)
return [p[1] for p in heap]
解法二:map+列表排序,时间O(nlogn),空间O(n)
import heapq
class Solution:
def topKFrequent(self, nums: List[int], k: int) -> List[int]:
#sol2
num2cnt = {}
for num in nums:
num2cnt[num] = num2cnt.get(num, 0) + 1
cnt2num = {}
for num, cnt in num2cnt.items():
try:
cnt2num[cnt].append(num)
except KeyError:
cnt2num[cnt] = [num]
cnts = sorted(cnt2num.keys())
res = []
idx = -1
while len(res) < k and idx >= -k:
res += cnt2num[cnts[idx]]
idx -= 1
return res