無秩序配列の中央値

無秩序配列の中央値
ソートアルゴリズムは使用できず,時間的複雑度O(n)が要求される.

# -*- coding: utf-8 -*-
​
# @Time    : 2019/4/22 10:42 AM
# @Author  : George
# @File    : main7.py
# @Contact : [email protected]
​
​
def heapify(seq, start, end):
    """
       start end        ，       
    :param seq:
    :param start:
    :param end:
    :return:
    """
    # start        
    left, right = 2 * start + 1, 2 * (start + 1)
    mi = start
    #             
    if left < end and seq[mi] > seq[left]:
        mi = left
    if right < end and seq[mi] > seq[right]:
        mi = right
    if mi != start:
        #           
        seq[start], seq[mi] = seq[mi], seq[start]
        heapify(seq, mi, end)
​
​
def find_mid_num(nums):
    heap_num = len(nums)//2
    heap = nums[:heap_num+1]
​
    #      
    start, end = len(heap) // 2 - 1, len(heap)
    for i in range(len(heap)//2-1, -1, -1):   #  n/2     
        heapify(heap, i, end)
​
    #                     ，        
    #             ，          
    for j in range(heap_num + 1, len(nums)):
        if nums[j] > heap[0]:
            #       
            print '  %s    %s' % (heap[0], nums[j])
            heap[0] = nums[j]
            heapify(heap, 0, end)
​
    #          ，            
    return heap[0] if len(nums) % 2 else (heap[0] + min(heap[1], heap[2])) / 2.0

転載先:https://www.cnblogs.com/George1994/p/10749180.html