如何生成列表的所有排列?
Posted
技术标签:
【中文标题】如何生成列表的所有排列?【英文标题】:How to generate all permutations of a list? 【发布时间】:2010-09-11 09:22:30 【问题描述】: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]
【问题讨论】:
我同意递归,接受的答案 - 今天。然而,这仍然是一个巨大的计算机科学问题。公认的答案以指数复杂度解决了这个问题 (2^N N=len(list)) 在多项式时间内解决它(或证明你不能):) 请参阅“旅行推销员问题” @FlipMcF 很难在多项式时间内“解决”它,因为它需要阶乘时间来枚举输出......所以,不,这是不可能的。 @FlipMcF: 不,这不是真的:a) 只是为了找到 最佳 解决方案,而不是 足够好 解决方案,这对于现实世界的目的和b)我们不需要扩展搜索空间中的所有节点,即所有排列;这就是heuristic algorithms like A* 【参考方案1】:def permutate(l):
for i, x in enumerate(l):
for y in l[i + 1:]:
yield x, y
if __name__ == '__main__':
print(list(permutate(list('abcd'))))
print(list(permutate([1, 2, 3, 4])))
#[('a', 'b'), ('a', 'c'), ('a', 'd'), ('b', 'c'), ('b', 'd'), ('c', 'd')]
#[(1, 2), (1, 3), (1, 4), (2, 3), (2, 4), (3, 4)]
【讨论】:
美观又简单!【参考方案2】: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") )
【讨论】:
【参考方案3】: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']
【讨论】:
【参考方案4】: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]]
【讨论】:
【参考方案5】:def permuteArray (arr):
arraySize = len(arr)
permutedList = []
if arraySize == 1:
return [arr]
i = 0
for item in arr:
for elem in permuteArray(arr[:i] + arr[i + 1:]):
permutedList.append([item] + elem)
i = i + 1
return permutedList
我不打算在新产品线中穷尽所有可能性,以使其有些独特。
【讨论】:
【参考方案6】:from typing import List
import time, random
def measure_time(func):
def wrapper_time(*args, **kwargs):
start_time = time.perf_counter()
res = func(*args, **kwargs)
end_time = time.perf_counter()
return res, end_time - start_time
return wrapper_time
class Solution:
def permute(self, nums: List[int], method: int = 1) -> List[List[int]]:
perms = []
perm = []
if method == 1:
_, time_perm = self._permute_recur(nums, 0, len(nums) - 1, perms)
elif method == 2:
_, time_perm = self._permute_recur_agian(nums, perm, perms)
print(perm)
return perms, time_perm
@measure_time
def _permute_recur(self, nums: List[int], l: int, r: int, perms: List[List[int]]):
# base case
if l == r:
perms.append(nums.copy())
for i in range(l, r + 1):
nums[l], nums[i] = nums[i], nums[l]
self._permute_recur(nums, l + 1, r , perms)
nums[l], nums[i] = nums[i], nums[l]
@measure_time
def _permute_recur_agian(self, nums: List[int], perm: List[int], perms_list: List[List[int]]):
"""
The idea is similar to nestedForLoops visualized as a recursion tree.
"""
if nums:
for i in range(len(nums)):
# perm.append(nums[i]) mistake, perm will be filled with all nums's elements.
# Method1 perm_copy = copy.deepcopy(perm)
# Method2 add in the parameter list using + (not in place)
# caveat: list.append is in-place , which is useful for operating on global element perms_list
# Note that:
# perms_list pass by reference. shallow copy
# perm + [nums[i]] pass by value instead of reference.
self._permute_recur_agian(nums[:i] + nums[i+1:], perm + [nums[i]], perms_list)
else:
# Arrive at the last loop, i.e. leaf of the recursion tree.
perms_list.append(perm)
if __name__ == "__main__":
array = [random.randint(-10, 10) for _ in range(3)]
sol = Solution()
# perms, time_perm = sol.permute(array, 1)
perms2, time_perm2 = sol.permute(array, 2)
print(perms2)
# print(perms, perms2)
# print(time_perm, time_perm2)
```
【讨论】:
一些解释会改善这个答案。【参考方案7】:如果有人喜欢这种丑陋的单线(虽然仅适用于字符串):
def p(a):
return a if len(a) == 1 else [[a[i], *j] for i in range(len(a)) for j in p(a[:i] + a[i + 1:])]
【讨论】:
【参考方案8】:使用递归求解,遍历元素,获取第 i 个元素,然后问自己:'其余元素的排列是什么',直到没有元素为止。
我在这里解释了解决方案:https://www.youtube.com/watch?v=_7GE7psS2b4
class Solution:
def permute(self,nums:List[int])->List[List[int]]:
res=[]
def dfs(nums,path):
if len(nums)==0:
res.append(path)
for i in range(len(nums)):
dfs(nums[:i]+nums[i+1:],path+[nums[i]])
dfs(nums,[])
return res
【讨论】:
【参考方案9】:对于 Python,我们可以使用 itertools 并导入排列和组合来解决您的问题
from itertools import product, permutations
A = ([1,2,3])
print (list(permutations(sorted(A),2)))
【讨论】:
以上是关于如何生成列表的所有排列?的主要内容,如果未能解决你的问题,请参考以下文章