如何生成列表的所有排列?
- 2024-11-19 08:39:00
- admin 原创
- 10
问题描述:
如何生成列表的所有排列?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]
解决方案 1:
itertools.permutations
从标准库中使用:
import itertools
list(itertools.permutations([1, 2, 3]))
itertools.permutations
如何实现的演示:
def permutations(elements):
if len(elements) <= 1:
yield elements
return
for perm in permutations(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替代方法itertools.permutations
。以下是其中一种:
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
另一个是基于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)
解决方案 2:
对于Python 2.6及以上版本:
import itertools
itertools.permutations([1, 2, 3])
这将返回一个生成器。用于list(permutations(xs))
返回一个列表。
解决方案 3:
首先,导入itertools
:
import itertools
排列(顺序很重要):
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)]
组合(顺序无关紧要):
print(list(itertools.combinations('123', 2)))
[('1', '2'), ('1', '3'), ('2', '3')]
笛卡尔积(具有多个可迭代项):
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)]
笛卡尔积(具有一个可迭代对象和其自身):
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)]
解决方案 4:
def permutations(head, tail=''):
if len(head) == 0:
print(tail)
else:
for i in range(len(head)):
permutations(head[:i] + head[i+1:], tail + head[i])
称为:
permutations('abc')
解决方案 5:
#!/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,需要可变序列类型作为输入。例如,perm(list("ball"))
可以工作,但perm("ball")
不能,因为您无法更改字符串。
这个 Python 实现的灵感来自于Horowitz、Sahni 和 Rajasekeran在《计算机算法》一书中提出的算法。
解决方案 6:
该解决方案实现了一个生成器,以避免将所有排列都保存在内存中:
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
解决方案 7:
常规实施(无收益 - 将在内存中完成所有操作):
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
收益率实现:
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第一个位置的所有元素,然后在第二个位置遍历除所选元素之外的所有其余元素,等等。您可以使用递归来执行此操作,其中停止标准是获取 1 个元素的数组 - 在这种情况下,您返回该数组。
解决方案 8:
功能性风格
def addperm(x,l):
return [ l[0:i] + [x] + l[i:] for i in range(len(l)+1) ]
def perm(l):
if len(l) == 0:
return [[]]
return [x for y in perm(l[1:]) for x in addperm(l[0],y) ]
print perm([ i for i in range(3)])
结果:
[[0, 1, 2], [1, 0, 2], [1, 2, 0], [0, 2, 1], [2, 0, 1], [2, 1, 0]]
解决方案 9:
以下代码是给定列表的就地排列,以生成器的形式实现。由于它仅返回对列表的引用,因此不应在生成器之外修改列表。该解决方案是非递归的,因此占用的内存较少。对于输入列表中元素的多个副本,也能很好地工作。
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
解决方案 10:
我认为一个相当明显的方法可能是:
def permutList(l):
if not l:
return [[]]
res = []
for e in l:
temp = l[:]
temp.remove(e)
res.extend([[e] + r for r in permutList(temp)])
return res
解决方案 11:
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]
]
解决方案 12:
我使用了一种基于阶乘数系统的算法- 对于长度为 n 的列表,您可以逐项组装每个排列,从每个阶段剩下的项目中进行选择。第一项有 n 个选择,第二项有 n-1 个选择,最后一项只有一个选择,因此您可以使用阶乘数系统中数字的数字作为索引。这样,数字 0 到 n!-1 就按字典顺序对应于所有可能的排列。
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]]
此方法是非递归的,但在我的计算机上速度稍慢,并且当 n! 太大而无法转换为 C 长整数(对我来说 n=13)时,xrange 会引发错误。当我需要它时,它就足够了,但它远远不如 itertools.permutations。
解决方案 13:
请注意,该算法的n factorial
时间复杂度为,其中n
是输入列表的长度
打印运行时的结果:
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()
例子:
permutation([1,2,3])
输出:
[1, 2, 3]
[1, 3, 2]
[2, 1, 3]
[2, 3, 1]
[3, 1, 2]
[3, 2, 1]
解决方案 14:
确实可以迭代每个排列的第一个元素,就像 tzwenn 的答案中那样。不过,这样写这个解决方案更有效:
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
这个解决方案大约快了 30%,显然这要归功于以 结尾的递归,len(elements) <= 1
而不是0
。它还更节省内存,因为它使用生成器函数(通过yield
),就像 Riccardo Reyes 的解决方案一样。
解决方案 15:
这是受到使用列表推导的 Haskell 实现的启发:
def permutation(list):
if len(list) == 0:
return [[]]
else:
return [[x] + ys for x in list for ys in permutation(delete(list, x))]
def delete(list, item):
lc = list[:]
lc.remove(item)
return lc
解决方案 16:
为了提高性能,受Knuth启发的 numpy 解决方案(p22):
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 - 比以下方法快 20 倍list(itertools.permutations(range(n))
:
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
解决方案 17:
如果您不想使用内置方法,例如:
import itertools
list(itertools.permutations([1, 2, 3]))
你可以自己实现排列函数
from collections.abc import Iterable
def permute(iterable: Iterable[str]) -> set[str]:
perms = set()
if len(iterable) == 1:
return {*iterable}
for index, char in enumerate(iterable):
perms.update([char + perm for perm in permute(iterable[:index] + iterable[index + 1:])])
return perms
if __name__ == '__main__':
print(permute('abc'))
# {'bca', 'abc', 'cab', 'acb', 'cba', 'bac'}
print(permute(['1', '2', '3']))
# {'123', '312', '132', '321', '213', '231'}
解决方案 18:
免责声明:软件包作者无耻地推销。:)
trotter包与大多数实现不同,它生成的伪列表实际上并不包含排列,而是描述排列与排序中相应位置之间的映射,从而可以处理非常大的排列“列表”,如此演示所示,它在包含字母表中所有字母排列的伪列表中执行相当即时的操作和查找,而不会使用比典型网页更多的内存或处理。
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]
解决方案 19:
递归之美:
>>> 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']
解决方案 20:
另一种方法(不使用库)
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']))
解决方案 21:
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)
解决方案 22:
这是一个对列表进行操作的算法,无需创建新的中间列表,类似于https://stackoverflow.com/a/108651/184528中 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代码:http: //repl.it/J9v
解决方案 23:
该算法是最有效的算法,它避免了递归调用中的数组传递和操作,适用于 Python 2、3:
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)
解决方案 24:
def pzip(c, seq):
result = []
for item in seq:
for i in range(len(item)+1):
result.append(item[i:]+c+item[:i])
return result
def perm(line):
seq = [c for c in line]
if len(seq) <=1 :
return seq
else:
return pzip(seq[0], perm(seq[1:]))
解决方案 25:
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)
解决方案 26:
我看到这些递归函数内部进行了很多迭代,并不完全是纯递归......
所以对于那些连一个循环都不能忍受的人来说,这是一个糟糕的、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])
解决方案 27:
为了节省大家搜索和实验的时间,这里提供了 Python 中的非递归排列解决方案,它也适用于 Numba(截至 v. 0.41):
@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]]
给出关于性能的印象:
%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,仅当您必须从 njitted 函数调用它时才使用此版本,否则最好使用 itertools 实现。
解决方案 28:
无论如何,我们可以使用sympy库,也支持多集排列
import sympy
from sympy.utilities.iterables import multiset_permutations
t = [1,2,3]
p = list(multiset_permutations(t))
print(p)
# [[1, 2, 3], [1, 3, 2], [2, 1, 3], [2, 3, 1], [3, 1, 2], [3, 2, 1]]
答案受到获取 numpy 数组的所有排列的极大启发
解决方案 29:
另一个解决方案:
def permutation(flag, k =1 ):
N = len(flag)
for i in xrange(0, N):
if flag[i] != 0:
continue
flag[i] = k
if k == N:
print flag
permutation(flag, k+1)
flag[i] = 0
permutation([0, 0, 0])
解决方案 30:
这是初始排序后生成排列的渐近最优方法 O(n*n!)。
最多有 n! 种排列,并且 hasNextPermutation(..) 的运行时间复杂度为 O(n)
分为 3 个步骤,
找到最大的 j,使得 a[j] 可以增加
将 a[j] 增加最小可行量
按照字典顺序找到扩展新 a[0..j] 的最小方法
'''
Lexicographic permutation generation
consider example array state of [1,5,6,4,3,2] for sorted [1,2,3,4,5,6]
after 56432(treat as number) ->nothing larger than 6432(using 6,4,3,2) beginning with 5
so 6 is next larger and 2345(least using numbers other than 6)
so [1, 6,2,3,4,5]
'''
def hasNextPermutation(array, len):
' Base Condition '
if(len ==1):
return False
'''
Set j = last-2 and find first j such that a[j] < a[j+1]
If no such j(j==-1) then we have visited all permutations
after this step a[j+1]>=..>=a[len-1] and a[j]<a[j+1]
a[j]=5 or j=1, 6>5>4>3>2
'''
j = len -2
while (j >= 0 and array[j] >= array[j + 1]):
j= j-1
if(j==-1):
return False
# print(f"After step 2 for j {j} {array}")
'''
decrease l (from n-1 to j) repeatedly until a[j]<a[l]
Then swap a[j], a[l]
a[l] is the smallest element > a[j] that can follow a[l]...a[j-1] in permutation
before swap we have a[j+1]>=..>=a[l-1]>=a[l]>a[j]>=a[l+1]>=..>=a[len-1]
after swap -> a[j+1]>=..>=a[l-1]>=a[j]>a[l]>=a[l+1]>=..>=a[len-1]
a[l]=6 or l=2, j=1 just before swap [1, 5, 6, 4, 3, 2]
after swap [1, 6, 5, 4, 3, 2] a[l]=5, a[j]=6
'''
l = len -1
while(array[j] >= array[l]):
l = l-1
# print(f"After step 3 for l={l}, j={j} before swap {array}")
array[j], array[l] = array[l], array[j]
# print(f"After step 3 for l={l} j={j} after swap {array}")
'''
Reverse a[j+1...len-1](both inclusive)
after reversing [1, 6, 2, 3, 4, 5]
'''
array[j+1:len] = reversed(array[j+1:len])
# print(f"After step 4 reversing {array}")
return True
array = [1,2,4,4,5]
array.sort()
len = len(array)
count =1
print(array)
'''
The algorithm visits every permutation in lexicographic order
generating one by one
'''
while(hasNextPermutation(array, len)):
print(array)
count = count +1
# The number of permutations will be n! if no duplicates are present, else less than that
# [1,4,3,3,2] -> 5!/2!=60
print(f"Number of permutations: {count}")
- 2024年20款好用的项目管理软件推荐,项目管理提效的20个工具和技巧
- 2024年开源项目管理软件有哪些?推荐5款好用的项目管理工具
- 项目管理软件有哪些?推荐7款超好用的项目管理工具
- 项目管理软件哪个最好用?盘点推荐5款好用的项目管理工具
- 项目管理软件有哪些最好用?推荐6款好用的项目管理工具
- 项目管理软件有哪些,盘点推荐国内外超好用的7款项目管理工具
- 2024项目管理软件排行榜(10类常用的项目管理工具全推荐)
- 项目管理软件排行榜:2024年项目经理必备5款开源项目管理软件汇总
- 2024年常用的项目管理软件有哪些?推荐这10款国内外好用的项目管理工具
- 项目管理必备:盘点2024年13款好用的项目管理软件