Top K问题用Python求解

用Python写的Top K问题，一种是对N个数据直接排序，然后取出其最大的K个数。复杂度为N*logN。

一种是通过保留K个数的槽，然后不断去N个数中取数，如果取出的数比K个槽中的数的最小值要大，那么就替换这个最小值。所以时间复杂度为N*logK.以下为时间。

复杂度	N*logk	N*logN
1W	0.0049101173134	0.00381023735085
10W	0.0173171225053	0.0491447920128
100W	0.125165655521	0.928009178781
1000W	1.51472073013	12.1190111173

K等于100

  
  

   
   

    
    
   
   
#coding=utf-8
'''
Created on Sep 21, 2015
@author: haoyifen
'''
import time
import random
def left(i):
    return 2*i+1
def right(i):
    return 2*i+2
def min_heapify(a,heap_size,i):
    l=left(i)
    r=right(i)
    if l<heap_size and a[l]<a[i] :
        least=l
    else:
        least=i
    if r<heap_size and  a[r]<a[least]:
        least=r
    if least!=i:
        a[i],a[least]=a[least],a[i]
        min_heapify(a, heap_size, least)
def build_min_heap(a,heap_size):
    for i in range((heap_size-1)/2,-1,-1):
        min_heapify(a, heap_size, i)
def heap_sort(a,heap_size):
    build_min_heap(a, heap_size)
    print a 
    for i in range(heap_size-1,0,-1):
        a[i],a[0]=a[0],a[i]
        #print a
        heap_size-=1
        min_heapify(a, heap_size, 0)
def findMaxKInN(N, K):
    All_Data = []
    for i in range(N):
        ele = random.randint(1, N)
        All_Data.append(ele)
    a = []
    for i in range(K):
        a.append(All_Data[i])
    start = time.clock()
    build_min_heap(a, len(a))
    for i in range(K, N, 1):
        if a[0] < All_Data[i]:
            a[0] = All_Data[i]
            min_heapify(a, K, 0)
    end = time.clock()
    print "N=",N,"K=",K
    print "N*logK" + "复杂度所用的时间", end - start
    start = time.clock()
    All_Data.sort(cmp=None, key=None, reverse=True)
    end = time.clock()
    print "使用排序N*logN", end - start
if __name__=="__main__":
    N=10000
    K=100
    while(N<=10000000):
        findMaxKInN(N, K)
        N*=10;