第k序元素查找python,有没有一种方法可以在python中对列表进行排序，直到找到第一个排序后的k个元素？...

您可以使用^{}模块，特别是它的^{}或{a3}函数。在

或者只需构建堆并调用^{}。这需要O(n)时间来构建堆，O(k*log(n))来检索k元素。在

这里有一个非常简单和小的基准：In [1]: import random, heapq

In [2]: seq = [random.randint(-5000, 5000) for _ in range(35000)]

In [3]: %timeit sorted(seq)[:75]

100 loops, best of 3: 14.5 ms per loop

In [4]: %%timeit

...: s = seq[:]

...: heapq.nsmallest(75, s)

...:

100 loops, best of 3: 4.05 ms per loop

In [5]: %%timeit

...: s = seq[:]

...: heapq.heapify(s)

...: for _ in range(75): heapq.heappop(s)

...:

100 loops, best of 3: 2.41 ms per loop

我不知道为什么nsmallest比直接调用heappop慢得多。实际上，我应该在不复制seq的情况下计时，但仍然：

^{pr2}$

长度增加100倍：In [12]: %timeit sorted(seq)[:75]

1 loops, best of 3: 1.9 s per loop

In [13]: %%timeit

...: heapq.nsmallest(75, seq)

...:

1 loops, best of 3: 352 ms per loop

In [14]: %%timeit

...: s = seq[:]

...: heapq.heapify(s)

...: for _ in range(75): heapq.heappop(s)

...:

1 loops, best of 3: 264 ms per loop

注：为了对抗F.J偏压分析：In [13]: a = list(range(1000000))

In [14]: random.shuffle(a)

In [15]: %timeit sorted(a)

1 loops, best of 3: 985 ms per loop

In [16]: %%timeit

...: s = a[:]

...: heapq.heapify(s)

...:

1 loops, best of 3: 284 ms per loop

如您所见，heapify甚至比在1000000个元素列表上排序要快得多。在