如何在Python中生成列表的所有排列,而与列表中元素的类型无关?
例如:
permutations([])
[]
permutations([1])
[1]
permutations([1, 2])
[1, 2]
[2, 1]
permutations([1, 2, 3])
[1, 2, 3]
[1, 3, 2]
[2, 1, 3]
[2, 3, 1]
[3, 1, 2]
[3, 2, 1]
如何在Python中生成列表的所有排列,而与列表中元素的类型无关?
例如:
permutations([])
[]
permutations([1])
[1]
permutations([1, 2])
[1, 2]
[2, 1]
permutations([1, 2, 3])
[1, 2, 3]
[1, 3, 2]
[2, 1, 3]
[2, 3, 1]
[3, 1, 2]
[3, 2, 1]
Answers:
从Python 2.6开始(如果您使用的是Python 3),您可以使用标准库工具:itertools.permutations
。
import itertools
list(itertools.permutations([1, 2, 3]))
如果您出于某种原因使用旧版Python(<2.6),或者只是想知道它的工作原理,那么这是一种不错的方法,摘自 http://code.activestate.com/recipes/252178/:
def all_perms(elements):
if len(elements) <=1:
yield elements
else:
for perm in all_perms(elements[1:]):
for i in range(len(elements)):
# nb elements[0:1] works in both string and list contexts
yield perm[:i] + elements[0:1] + perm[i:]
的文档中列出了几种其他方法itertools.permutations
。这是一个:
def permutations(iterable, r=None):
# permutations('ABCD', 2) --> AB AC AD BA BC BD CA CB CD DA DB DC
# permutations(range(3)) --> 012 021 102 120 201 210
pool = tuple(iterable)
n = len(pool)
r = n if r is None else r
if r > n:
return
indices = range(n)
cycles = range(n, n-r, -1)
yield tuple(pool[i] for i in indices[:r])
while n:
for i in reversed(range(r)):
cycles[i] -= 1
if cycles[i] == 0:
indices[i:] = indices[i+1:] + indices[i:i+1]
cycles[i] = n - i
else:
j = cycles[i]
indices[i], indices[-j] = indices[-j], indices[i]
yield tuple(pool[i] for i in indices[:r])
break
else:
return
另一个基于itertools.product
:
def permutations(iterable, r=None):
pool = tuple(iterable)
n = len(pool)
r = n if r is None else r
for indices in product(range(n), repeat=r):
if len(set(indices)) == r:
yield tuple(pool[i] for i in indices)
for i in range(len(elements))
而不是for i in range(len(elements)+1)
。实际上,单选元素elements[0:1]
可以位于len(elements)
不同位置,结果不是len(elements)+1
。
在Python 2.6及更高版本中:
import itertools
itertools.permutations([1,2,3])
(作为生成器返回。用于list(permutations(l))
作为列表返回。)
r
参数,例如itertools.permutations([1,2,3], r=2)
,它将生成选择2个元素的所有可能的排列:[(1, 2), (1, 3), (2, 1), (2, 3), (3, 1), (3, 2)]
以下代码仅适用于Python 2.6及更高版本
首先,导入itertools
:
import itertools
print list(itertools.permutations([1,2,3,4], 2))
[(1, 2), (1, 3), (1, 4),
(2, 1), (2, 3), (2, 4),
(3, 1), (3, 2), (3, 4),
(4, 1), (4, 2), (4, 3)]
print list(itertools.combinations('123', 2))
[('1', '2'), ('1', '3'), ('2', '3')]
print list(itertools.product([1,2,3], [4,5,6]))
[(1, 4), (1, 5), (1, 6),
(2, 4), (2, 5), (2, 6),
(3, 4), (3, 5), (3, 6)]
print list(itertools.product([1,2], repeat=3))
[(1, 1, 1), (1, 1, 2), (1, 2, 1), (1, 2, 2),
(2, 1, 1), (2, 1, 2), (2, 2, 1), (2, 2, 2)]
def permutations(head, tail=''):
if len(head) == 0: print tail
else:
for i in range(len(head)):
permutations(head[0:i] + head[i+1:], tail+head[i])
称为:
permutations('abc')
#!/usr/bin/env python
def perm(a, k=0):
if k == len(a):
print a
else:
for i in xrange(k, len(a)):
a[k], a[i] = a[i] ,a[k]
perm(a, k+1)
a[k], a[i] = a[i], a[k]
perm([1,2,3])
输出:
[1, 2, 3]
[1, 3, 2]
[2, 1, 3]
[2, 3, 1]
[3, 2, 1]
[3, 1, 2]
当我交换列表的内容时,需要一个可变的序列类型作为输入。例如,perm(list("ball"))
将工作,perm("ball")
不会因为您不能更改字符串。
此Python实现受Horowitz,Sahni和Rajasekeran的《计算机算法》一书中介绍的算法的启发。
此解决方案实现了一个生成器,以避免将所有排列保留在内存中:
def permutations (orig_list):
if not isinstance(orig_list, list):
orig_list = list(orig_list)
yield orig_list
if len(orig_list) == 1:
return
for n in sorted(orig_list):
new_list = orig_list[:]
pos = new_list.index(n)
del(new_list[pos])
new_list.insert(0, n)
for resto in permutations(new_list[1:]):
if new_list[:1] + resto <> orig_list:
yield new_list[:1] + resto
以下代码是给定列表的就地排列,实现为生成器。由于仅返回对列表的引用,因此不应在生成器外部修改列表。该解决方案是非递归的,因此使用低内存。输入列表中元素的多个副本也可以很好地工作。
def permute_in_place(a):
a.sort()
yield list(a)
if len(a) <= 1:
return
first = 0
last = len(a)
while 1:
i = last - 1
while 1:
i = i - 1
if a[i] < a[i+1]:
j = last - 1
while not (a[i] < a[j]):
j = j - 1
a[i], a[j] = a[j], a[i] # swap the values
r = a[i+1:last]
r.reverse()
a[i+1:last] = r
yield list(a)
break
if i == first:
a.reverse()
return
if __name__ == '__main__':
for n in range(5):
for a in permute_in_place(range(1, n+1)):
print a
print
for a in permute_in_place([0, 0, 1, 1, 1]):
print a
print
list2Perm = [1, 2.0, 'three']
listPerm = [[a, b, c]
for a in list2Perm
for b in list2Perm
for c in list2Perm
if ( a != b and b != c and a != c )
]
print listPerm
输出:
[
[1, 2.0, 'three'],
[1, 'three', 2.0],
[2.0, 1, 'three'],
[2.0, 'three', 1],
['three', 1, 2.0],
['three', 2.0, 1]
]
我使用了基于阶乘数系统的算法-对于长度为n的列表,您可以逐项组合每个排列项,并从每个阶段剩下的项中进行选择。第一项有n个选择,第二项有n-1个,最后一项只有n个,因此可以将阶乘数字系统中数字的数字用作索引。这样,数字0到n!-1对应于字典顺序中所有可能的排列。
from math import factorial
def permutations(l):
permutations=[]
length=len(l)
for x in xrange(factorial(length)):
available=list(l)
newPermutation=[]
for radix in xrange(length, 0, -1):
placeValue=factorial(radix-1)
index=x/placeValue
newPermutation.append(available.pop(index))
x-=index*placeValue
permutations.append(newPermutation)
return permutations
permutations(range(3))
输出:
[[0, 1, 2], [0, 2, 1], [1, 0, 2], [1, 2, 0], [2, 0, 1], [2, 1, 0]]
该方法是非递归的,但是在我的计算机上它会稍微慢一些,并且当n!时xrange会引发错误!太大而无法转换为C长整数(对我来说n = 13)。当我需要它时就足够了,但是从长远来看这不是itertools.permutations。
请注意,此算法具有n factorial
时间复杂度,其中n
是输入列表的长度
打印运行结果:
global result
result = []
def permutation(li):
if li == [] or li == None:
return
if len(li) == 1:
result.append(li[0])
print result
result.pop()
return
for i in range(0,len(li)):
result.append(li[i])
permutation(li[:i] + li[i+1:])
result.pop()
例:
permutation([1,2,3])
输出:
[1, 2, 3]
[1, 3, 2]
[2, 1, 3]
[2, 3, 1]
[3, 1, 2]
[3, 2, 1]
正如tzwenn的回答,确实可以迭代每个排列的第一个元素。但是,以这种方式编写此解决方案效率更高:
def all_perms(elements):
if len(elements) <= 1:
yield elements # Only permutation possible = no permutation
else:
# Iteration over the first element in the result permutation:
for (index, first_elmt) in enumerate(elements):
other_elmts = elements[:index]+elements[index+1:]
for permutation in all_perms(other_elmts):
yield [first_elmt] + permutation
该解决方案的速度提高了约30%,这显然归功于递归以len(elements) <= 1
代替0
。yield
就像Riccardo Reyes的解决方案一样,它使用生成器函数(通过),因此内存效率也更高。
常规执行(无收益-将在内存中做所有事情):
def getPermutations(array):
if len(array) == 1:
return [array]
permutations = []
for i in range(len(array)):
# get all perm's of subarray w/o current item
perms = getPermutations(array[:i] + array[i+1:])
for p in perms:
permutations.append([array[i], *p])
return permutations
收益实施:
def getPermutations(array):
if len(array) == 1:
yield array
else:
for i in range(len(array)):
perms = getPermutations(array[:i] + array[i+1:])
for p in perms:
yield [array[i], *p]
基本思想是遍历数组中所有元素的第1个位置,然后在第2个位置中遍历所有其余元素,而第1个位置没有选择元素,依此类推。您可以使用recursion进行操作,其中停止条件是到达由1个元素组成的数组-在这种情况下,您将返回该数组。
perms = getPermutations(array[:i] + array[i+1:])
numpy
数组_> getPermutations(np.array([1, 2, 3]))
,我看到它适用于列表,只是因为func arg感到困惑array
:)
numba
并且对速度感到贪婪,因此试图仅将其与numpy
数组一起使用
为了提高性能,从Knuth(p22)那里得到了一个麻木的解决方案:
from numpy import empty, uint8
from math import factorial
def perms(n):
f = 1
p = empty((2*n-1, factorial(n)), uint8)
for i in range(n):
p[i, :f] = i
p[i+1:2*i+1, :f] = p[:i, :f] # constitution de blocs
for j in range(i):
p[:i+1, f*(j+1):f*(j+2)] = p[j+1:j+i+2, :f] # copie de blocs
f = f*(i+1)
return p[:n, :]
复制大块内存可以节省时间-比list(itertools.permutations(range(n))
以下方法快20倍:
In [1]: %timeit -n10 list(permutations(range(10)))
10 loops, best of 3: 815 ms per loop
In [2]: %timeit -n100 perms(10)
100 loops, best of 3: 40 ms per loop
from __future__ import print_function
def perm(n):
p = []
for i in range(0,n+1):
p.append(i)
while True:
for i in range(1,n+1):
print(p[i], end=' ')
print("")
i = n - 1
found = 0
while (not found and i>0):
if p[i]<p[i+1]:
found = 1
else:
i = i - 1
k = n
while p[i]>p[k]:
k = k - 1
aux = p[i]
p[i] = p[k]
p[k] = aux
for j in range(1,(n-i)/2+1):
aux = p[i+j]
p[i+j] = p[n-j+1]
p[n-j+1] = aux
if not found:
break
perm(5)
这是一种适用于列表的算法,无需创建新的中间列表,类似于https://stackoverflow.com/a/108651/184528上Ber的解决方案。
def permute(xs, low=0):
if low + 1 >= len(xs):
yield xs
else:
for p in permute(xs, low + 1):
yield p
for i in range(low + 1, len(xs)):
xs[low], xs[i] = xs[i], xs[low]
for p in permute(xs, low + 1):
yield p
xs[low], xs[i] = xs[i], xs[low]
for p in permute([1, 2, 3, 4]):
print p
您可以在此处亲自尝试该代码:http : //repl.it/J9v
递归的美丽:
>>> import copy
>>> def perm(prefix,rest):
... for e in rest:
... new_rest=copy.copy(rest)
... new_prefix=copy.copy(prefix)
... new_prefix.append(e)
... new_rest.remove(e)
... if len(new_rest) == 0:
... print new_prefix + new_rest
... continue
... perm(new_prefix,new_rest)
...
>>> perm([],['a','b','c','d'])
['a', 'b', 'c', 'd']
['a', 'b', 'd', 'c']
['a', 'c', 'b', 'd']
['a', 'c', 'd', 'b']
['a', 'd', 'b', 'c']
['a', 'd', 'c', 'b']
['b', 'a', 'c', 'd']
['b', 'a', 'd', 'c']
['b', 'c', 'a', 'd']
['b', 'c', 'd', 'a']
['b', 'd', 'a', 'c']
['b', 'd', 'c', 'a']
['c', 'a', 'b', 'd']
['c', 'a', 'd', 'b']
['c', 'b', 'a', 'd']
['c', 'b', 'd', 'a']
['c', 'd', 'a', 'b']
['c', 'd', 'b', 'a']
['d', 'a', 'b', 'c']
['d', 'a', 'c', 'b']
['d', 'b', 'a', 'c']
['d', 'b', 'c', 'a']
['d', 'c', 'a', 'b']
['d', 'c', 'b', 'a']
此算法是最有效的算法,它避免了在递归调用中进行数组传递和操作,在Python 2、3中有效:
def permute(items):
length = len(items)
def inner(ix=[]):
do_yield = len(ix) == length - 1
for i in range(0, length):
if i in ix: #avoid duplicates
continue
if do_yield:
yield tuple([items[y] for y in ix + [i]])
else:
for p in inner(ix + [i]):
yield p
return inner()
用法:
for p in permute((1,2,3)):
print(p)
(1, 2, 3)
(1, 3, 2)
(2, 1, 3)
(2, 3, 1)
(3, 1, 2)
(3, 2, 1)
另一种方法(无库)
def permutation(input):
if len(input) == 1:
return input if isinstance(input, list) else [input]
result = []
for i in range(len(input)):
first = input[i]
rest = input[:i] + input[i + 1:]
rest_permutation = permutation(rest)
for p in rest_permutation:
result.append(first + p)
return result
输入可以是字符串或列表
print(permutation('abcd'))
print(permutation(['a', 'b', 'c', 'd']))
[1, 2, 3]
返回[6, 6, 6, 6, 6, 6]
print(permutation(['1','2','3']))
免责声明:软件包作者提供的无形插件。:)
该猪手包是从大多数实现不同,它产生不实际包含的排列,而是描述的排列和各位置之间的映射关系的顺序,使其能够工作,排列非常大名单“,如图所示伪名单在这个演示中,执行了一个非常瞬时的操作并在“包含”字母中所有字母排列的伪列表中进行查找,而没有使用比典型的Web页面更多的内存或处理。
无论如何,要生成排列列表,我们可以执行以下操作。
import trotter
my_permutations = trotter.Permutations(3, [1, 2, 3])
print(my_permutations)
for p in my_permutations:
print(p)
输出:
包含[1、2、3]的6个3排列的伪列表。 [1,2,3] [1、3、2] [3,1,2] [3,2,1] [2,3,1] [2,1,3]
生成所有可能的排列
我正在使用python3.4:
def calcperm(arr, size):
result = set([()])
for dummy_idx in range(size):
temp = set()
for dummy_lst in result:
for dummy_outcome in arr:
if dummy_outcome not in dummy_lst:
new_seq = list(dummy_lst)
new_seq.append(dummy_outcome)
temp.add(tuple(new_seq))
result = temp
return result
测试用例:
lst = [1, 2, 3, 4]
#lst = ["yellow", "magenta", "white", "blue"]
seq = 2
final = calcperm(lst, seq)
print(len(final))
print(final)
为了节省大家的搜索和实验时间,以下是Python中的非递归置换解决方案,该解决方案也适用于Numba(自0.41版起):
@numba.njit()
def permutations(A, k):
r = [[i for i in range(0)]]
for i in range(k):
r = [[a] + b for a in A for b in r if (a in b)==False]
return r
permutations([1,2,3],3)
[[1, 2, 3], [1, 3, 2], [2, 1, 3], [2, 3, 1], [3, 1, 2], [3, 2, 1]]
给人印象的表现:
%timeit permutations(np.arange(5),5)
243 µs ± 11.1 µs per loop (mean ± std. dev. of 7 runs, 1 loop each)
time: 406 ms
%timeit list(itertools.permutations(np.arange(5),5))
15.9 µs ± 8.61 ns per loop (mean ± std. dev. of 7 runs, 100000 loops each)
time: 12.9 s
因此,仅当必须从njitted函数调用它时才使用此版本,否则,请选择itertools实现。
我看到这些递归函数内部进行了很多迭代,而并非完全是纯函数递归...
因此对于那些甚至无法遵守一个循环的人来说,这是一个总的,完全不必要的完全递归解决方案
def all_insert(x, e, i=0):
return [x[0:i]+[e]+x[i:]] + all_insert(x,e,i+1) if i<len(x)+1 else []
def for_each(X, e):
return all_insert(X[0], e) + for_each(X[1:],e) if X else []
def permute(x):
return [x] if len(x) < 2 else for_each( permute(x[1:]) , x[0])
perms = permute([1,2,3])
我的Python解决方案:
def permutes(input,offset):
if( len(input) == offset ):
return [''.join(input)]
result=[]
for i in range( offset, len(input) ):
input[offset], input[i] = input[i], input[offset]
result = result + permutes(input,offset+1)
input[offset], input[i] = input[i], input[offset]
return result
# input is a "string"
# return value is a list of strings
def permutations(input):
return permutes( list(input), 0 )
# Main Program
print( permutations("wxyz") )
def permutation(word, first_char=None):
if word == None or len(word) == 0: return []
if len(word) == 1: return [word]
result = []
first_char = word[0]
for sub_word in permutation(word[1:], first_char):
result += insert(first_char, sub_word)
return sorted(result)
def insert(ch, sub_word):
arr = [ch + sub_word]
for i in range(len(sub_word)):
arr.append(sub_word[i:] + ch + sub_word[:i])
return arr
assert permutation(None) == []
assert permutation('') == []
assert permutation('1') == ['1']
assert permutation('12') == ['12', '21']
print permutation('abc')
输出:['abc','acb','bac','bca','cab','cba']
使用 Counter
from collections import Counter
def permutations(nums):
ans = [[]]
cache = Counter(nums)
for idx, x in enumerate(nums):
result = []
for items in ans:
cache1 = Counter(items)
for id, n in enumerate(nums):
if cache[n] != cache1[n] and items + [n] not in result:
result.append(items + [n])
ans = result
return ans
permutations([1, 2, 2])
> [[1, 2, 2], [2, 1, 2], [2, 2, 1]]