|
| 1 | +#python中的堆排序peapq模块 |
| 2 | + |
| 3 | +heapq模块实现了python中的堆排序,并提供了有关方法。让用Python实现排序算法有了简单快捷的方式。 |
| 4 | + |
| 5 | +heapq的官方文档和源码:[8.4.heapq-Heap queue algorithm](https://docs.python.org/2/library/heapq.html) |
| 6 | + |
| 7 | +下面通过举例的方式说明heapq的应用方法 |
| 8 | + |
| 9 | +##实现堆排序 |
| 10 | + |
| 11 | + #! /usr/bin/evn python |
| 12 | + #coding:utf-8 |
| 13 | + |
| 14 | + from heapq import * |
| 15 | + |
| 16 | + def heapsort(iterable): |
| 17 | + h = [] |
| 18 | + for value in iterable: |
| 19 | + heappush(h,value) |
| 20 | + return [heappop(h) for i in range(len(h))] |
| 21 | + |
| 22 | + if __name__=="__main__": |
| 23 | + print heapsort([1,3,5,9,2]) |
| 24 | + |
| 25 | +###heappush() |
| 26 | + |
| 27 | +heapq.heappush(heap, item):将item压入到堆数组heap中。如果不进行此步操作,后面的heappop()失效 |
| 28 | + |
| 29 | +###heappop() |
| 30 | + |
| 31 | +heapq.heappop(heap):从堆数组heap中取出最小的值,并返回。 |
| 32 | + |
| 33 | + >>> h=[] #定义一个list |
| 34 | + >>> from heapq import * #引入heapq模块 |
| 35 | + >>> h |
| 36 | + [] |
| 37 | + >>> heappush(h,5) #向堆中依次增加数值 |
| 38 | + >>> heappush(h,2) |
| 39 | + >>> heappush(h,3) |
| 40 | + >>> heappush(h,9) |
| 41 | + >>> h #h的值 |
| 42 | + [2, 5, 3, 9] |
| 43 | + >>> heappop(h) #从h中删除最小的,并返回该值 |
| 44 | + 2 |
| 45 | + >>> h |
| 46 | + [3, 5, 9] |
| 47 | + >>> h.append(1) #注意,如果不是压入堆中,而是通过append追加一个数值 |
| 48 | + >>> h #堆的函数并不能操作这个增加的数值,或者说它堆对来讲是不存在的 |
| 49 | + [3, 5, 9, 1] |
| 50 | + >>> heappop(h) #从h中能够找到的最小值是3,而不是1 |
| 51 | + 3 |
| 52 | + >>> heappush(h,2) #这时,不仅将2压入到堆内,而且1也进入了堆。 |
| 53 | + >>> h |
| 54 | + [1, 2, 9, 5] |
| 55 | + >>> heappop(h) #操作对象已经包含了1 |
| 56 | + 1 |
| 57 | + |
| 58 | +###heapq.heappushpop(heap, item) |
| 59 | + |
| 60 | +是上述heappush和heappop的合体,同时完成两者的功能.注意:相当于先操作了heappush(heap,item),然后操作heappop(heap) |
| 61 | + |
| 62 | + >>> h |
| 63 | + [1, 2, 9, 5] |
| 64 | + >>> heappop(h) |
| 65 | + 1 |
| 66 | + >>> heappushpop(h,4) #增加4同时删除最小值2并返回该最小值,与下列操作等同: |
| 67 | + 2 #heappush(h,4),heappop(h) |
| 68 | + >>> h |
| 69 | + [4, 5, 9] |
| 70 | + |
| 71 | +###heapq.heapify(x) |
| 72 | + |
| 73 | +x必须是list,此函数将list变成堆,实时操作。从而能够在任何情况下使用堆的函数。 |
| 74 | + |
| 75 | + >>> a=[3,6,1] |
| 76 | + >>> heapify(a) #将a变成堆之后,可以对其操作 |
| 77 | + >>> heappop(a) |
| 78 | + 1 |
| 79 | + >>> b=[4,2,5] #b不是堆,如果对其进行操作,显示结果如下 |
| 80 | + >>> heappop(b) #按照顺序,删除第一个数值并返回,不会从中挑选出最小的 |
| 81 | + 4 |
| 82 | + >>> heapify(b) #变成堆之后,再操作 |
| 83 | + >>> heappop(b) |
| 84 | + 2 |
| 85 | + |
| 86 | +###heapq.heapreplace(heap, item) |
| 87 | + |
| 88 | +是heappop(heap)和heappush(heap,item)的联合操作。注意,与heappushpop(heap,item)的区别在于,顺序不同,这里是先进行删除,后压入堆 |
| 89 | + |
| 90 | + >>> a=[] |
| 91 | + >>> heapreplace(a,3) #如果list空,则报错 |
| 92 | + Traceback (most recent call last): |
| 93 | + File "<stdin>", line 1, in <module> |
| 94 | + IndexError: index out of range |
| 95 | + >>> heappush(a,3) |
| 96 | + >>> a |
| 97 | + [3] |
| 98 | + >>> heapreplace(a,2) #先执行删除(heappop(a)->3),再执行加入(heappush(a,2)) |
| 99 | + 3 |
| 100 | + >>> a |
| 101 | + [2] |
| 102 | + >>> heappush(a,5) |
| 103 | + >>> heappush(a,9) |
| 104 | + >>> heappush(a,4) |
| 105 | + >>> a |
| 106 | + [2, 4, 9, 5] |
| 107 | + >>> heapreplace(a,6) #先从堆a中找出最小值并返回,然后加入6 |
| 108 | + 2 |
| 109 | + >>> a |
| 110 | + [4, 5, 9, 6] |
| 111 | + >>> heapreplace(a,1) #1是后来加入的,在1加入之前,a中的最小值是4 |
| 112 | + 4 |
| 113 | + >>> a |
| 114 | + [1, 5, 9, 6] |
| 115 | + |
| 116 | +###heapq.merge(\*iterables) |
| 117 | + |
| 118 | +举例: |
| 119 | + |
| 120 | + >>> a=[2,4,6] |
| 121 | + >>> b=[1,3,5] |
| 122 | + >>> c=merge(a,b) |
| 123 | + >>> list(c) |
| 124 | + [1, 2, 3, 4, 5, 6] |
| 125 | + |
| 126 | +在[归并排序](https://github.com/qiwsir/algorithm/blob/master/merge_sort.md)中详细演示了本函数的使用方法。 |
| 127 | + |
| 128 | +###heapq.nlargest(n, iterable[, key]),heapq.nsmallest(n, iterable[, key]) |
| 129 | + |
| 130 | +获取列表中最大、最小的几个值。 |
| 131 | + |
| 132 | + >>> a |
| 133 | + [2, 4, 6] |
| 134 | + >>> nlargest(2,a) |
| 135 | + [6, 4] |
| 136 | + |
0 commit comments