如何生成列表的所有排列?

发布于 2024-07-05 06:26:33 字数 213 浏览 8 评论 0原文

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]

How do I generate all the permutations of a list? For example:

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]

如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。

扫码二维码加入Web技术交流群

发布评论

需要 登录 才能够评论, 你可以免费 注册 一个本站的账号。

评论(30

青芜 2024-07-12 06:26:35

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']

The beauty of recursion:

>>> 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']
山人契 2024-07-12 06:26:35

另一种方法(没有库)

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']))

ANOTHER APPROACH (without libs)

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

Input can be a string or a list

print(permutation('abcd'))
print(permutation(['a', 'b', 'c', 'd']))
潦草背影 2024-07-12 06:26:35
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)
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)
故乡的云 2024-07-12 06:26:35

这是初始排序后生成排列的渐近最优方式 O(n*n!)。

有n个! 最多排列,hasNextPermutation(..) 的运行时间复杂度为 O(n)

分 3 步,

  1. 找到最大的 j,使得 a[j] 可以增加
  2. 将 a[j] 增加最小可行量
  3. 找到字典上最少的方法扩展新的 a[0..j]
'''
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}")


This is the asymptotically optimal way O(n*n!) of generating permutations after initial sorting.

There are n! permutations at most and hasNextPermutation(..) runs in O(n) time complexity

In 3 steps,

  1. Find largest j such that a[j] can be increased
  2. Increase a[j] by smallest feasible amount
  3. Find lexicogrpahically least way to extend the new a[0..j]
'''
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}")


笑看君怀她人 2024-07-12 06:26:35

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]]

答案深受 获取 numpy 数组的所有排列

Anyway we could use sympy library , also support for multiset permutations

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]]

Answer is highly inspired by Get all permutations of a numpy array

悲歌长辞 2024-07-12 06:26:35

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])

Another solution:

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])
℡寂寞咖啡 2024-07-12 06:26:35

为了节省大家可能的搜索和实验时间,这里是 Python 中的非递归排列解决方案,它也适用于 Numba(从 v. 0.41 开始):

@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]]

为了给大家留下关于性能的印象:

%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,仅当您必须调用时才使用此版本它来自 njitted 函数,否则更喜欢 itertools 实现。

To save you folks possible hours of searching and experimenting, here's the non-recursive permutaions solution in Python which also works with Numba (as of 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]]

To give an impression about performance:

%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

So use this version only if you have to call it from njitted function, otherwise prefer itertools implementation.

莫相离 2024-07-12 06:26:35

我看到这些递归函数内部发生了很多迭代,而不是完全纯粹递归......

所以对于那些甚至无法遵守单个循环的人来说,这里有一个粗鲁的、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])

I see a lot of iteration going on inside these recursive functions, not exactly pure recursion...

so for those of you who cannot abide by even a single loop, here's a gross, totally unnecessary fully recursive solution

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])
殊姿 2024-07-12 06:26:35
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:]))
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:]))
七颜 2024-07-12 06:26:35

生成

我正在使用 python3.4 的所有可能的排列:

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)

Generate all possible permutations

I'm using 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

Test Cases:

lst = [1, 2, 3, 4]
#lst = ["yellow", "magenta", "white", "blue"]
seq = 2
final = calcperm(lst, seq)
print(len(final))
print(final)
横笛休吹塞上声 2024-07-12 06:26:35

这是一种在列表上运行的算法,无需创建新的中间列表,类似于 https://stackoverflow.com/a/108651 中 Ber 的解决方案/184528

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

您可以在这里亲自尝试代码:http://repl.it/J9v

Here is an algorithm that works on a list without creating new intermediate lists similar to Ber's solution at https://stackoverflow.com/a/108651/184528.

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

You can try the code out for yourself here: http://repl.it/J9v

网名女生简单气质 2024-07-12 06:26:35

该算法是最有效的一种,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)

This algorithm is the most effective one, it avoids of array passing and manipulation in recursive calls, works in 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()

Usage:

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)
妄司 2024-07-12 06:26:35

如果你不想使用内置方法,例如:

import itertools
list(itertools.permutations([1, 2, 3]))

你可以自己实现 permute 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'}

If you don't want to use the builtin methods such as:

import itertools
list(itertools.permutations([1, 2, 3]))

you can implement permute function yourself

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'}
一曲琵琶半遮面シ 2024-07-12 06:26:35

免责声明:软件包作者无耻的插件。 :)

trotter 包与大多数实现不同,因为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 这个演示,它在“包含”字母表中字母的所有排列的伪列表中执行非常即时的操作和查找,而不使用比典型网络更多的内存或处理页。

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:

A pseudo-list containing 6 3-permutations of [1, 2, 3].
[1, 2, 3]
[1, 3, 2]
[3, 1, 2]
[3, 2, 1]
[2, 3, 1]
[2, 1, 3]

Disclaimer: shameless plug by package author. :)

The trotter package is different from most implementations in that it generates pseudo lists that don't actually contain permutations but rather describe mappings between permutations and respective positions in an ordering, making it possible to work with very large 'lists' of permutations, as shown in this demo which performs pretty instantaneous operations and look-ups in a pseudo-list 'containing' all the permutations of the letters in the alphabet, without using more memory or processing than a typical web page.

In any case, to generate a list of permutations, we can do the following.

import trotter

my_permutations = trotter.Permutations(3, [1, 2, 3])

print(my_permutations)

for p in my_permutations:
    print(p)

Output:

A pseudo-list containing 6 3-permutations of [1, 2, 3].
[1, 2, 3]
[1, 3, 2]
[3, 1, 2]
[3, 2, 1]
[2, 3, 1]
[2, 1, 3]
独守阴晴ぅ圆缺 2024-07-12 06:26:35

为了性能,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 -
它比 list(itertools.permutations(range(n)) 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

For performance, a numpy solution inspired by Knuth, (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, :]

Copying large blocs of memory saves time -
it's 20x faster than 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
空城仅有旧梦在 2024-07-12 06:26:35

这是受到使用列表理解的 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

This is inspired by the Haskell implementation using list comprehension:

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
北方的巷 2024-07-12 06:26:35

人们确实可以迭代每个排列的第一个元素,就像 tzwenn 的答案一样。 然而,以这种方式编写此解决方案的效率更高:

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

此解决方案的速度大约快 30%,这显然要归功于以 len(elements) <= 1 结束的递归,而不是 0< /代码>。
它的内存效率也更高,因为它使用生成器函数(通过 yield),就像 Riccardo Reyes 的解决方案一样。

One can indeed iterate over the first element of each permutation, as in tzwenn's answer. It is however more efficient to write this solution this way:

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

This solution is about 30 % faster, apparently thanks to the recursion ending at len(elements) <= 1 instead of 0.
It is also much more memory-efficient, as it uses a generator function (through yield), like in Riccardo Reyes's solution.

弃爱 2024-07-12 06:26:34

请注意,该算法的时间复杂度为 n 阶乘,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()    

示例:

permutation([1,2,3])

输出:

[1, 2, 3]
[1, 3, 2]
[2, 1, 3]
[2, 3, 1]
[3, 1, 2]
[3, 2, 1]

Note that this algorithm has an n factorial time complexity, where n is the length of the input list

Print the results on the run:

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()    

Example:

permutation([1,2,3])

Output:

[1, 2, 3]
[1, 3, 2]
[2, 1, 3]
[2, 3, 1]
[3, 1, 2]
[3, 2, 1]
望笑 2024-07-12 06:26:33

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]]

In a functional style

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)])

The result:

[[0, 1, 2], [1, 0, 2], [1, 2, 0], [0, 2, 1], [2, 0, 1], [2, 1, 0]]
久隐师 2024-07-12 06:26:33

常规实现(无yield - 将在内存中执行所有操作):

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

yield 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所有其余元素,而无需选择第一个元素,等等。您可以使用递归来执行此操作,其中停止条件是获取包含 1 个元素的数组 - 在这种情况下您将返回该数组。

输入图像描述这里

Regular implementation (no yield - will do everything in memory):

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

Yield implementation:

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]

The basic idea is to go over all the elements in the array for the 1st position, and then in 2nd position go over all the rest of the elements without the chosen element for the 1st, etc. You can do this with recursion, where the stop criteria is getting to an array of 1 element - in which case you return that array.

enter image description here

那伤。 2024-07-12 06:26:33

该解决方案实现了一个生成器,以避免将所有排列保存在内存中:

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

This solution implements a generator, to avoid holding all the permutations on memory:

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
嘿哥们儿 2024-07-12 06:26:33
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')
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])

called as:

permutations('abc')
〃安静 2024-07-12 06:26:33
#!/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,需要一个可变序列类型作为输入。 例如, perm(list("ball")) 可以工作,而 perm("ball") 则不行,因为您无法更改字符串。

此 Python 实现的灵感来自于 Horowitz、Sahni 和 Rajasekeran 所著的《Computer Algorithms》一书中提出的算法。

#!/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])

Output:

[1, 2, 3]
[1, 3, 2]
[2, 1, 3]
[2, 3, 1]
[3, 2, 1]
[3, 1, 2]

As I'm swapping the content of the list it's required a mutable sequence type as input. E.g. perm(list("ball")) will work and perm("ball") won't because you can't change a string.

This Python implementation is inspired by the algorithm presented in the book Computer Algorithms by Horowitz, Sahni and Rajasekeran.

倚栏听风 2024-07-12 06:26:33

对于 Python 2.6 及以上:

import itertools
itertools.permutations([1, 2, 3])

这将作为生成器返回。 使用 list(permutations(xs)) 以列表形式返回。

For Python 2.6 onwards:

import itertools
itertools.permutations([1, 2, 3])

This returns as a generator. Use list(permutations(xs)) to return as a list.

情释 2024-07-12 06:26:33

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(具有多个可迭代对象):

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及其自身):

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)]

First, import itertools:

import itertools

Permutation (order matters):

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)]

Combination (order does NOT matter):

print(list(itertools.combinations('123', 2)))

[('1', '2'), ('1', '3'), ('2', '3')]

Cartesian product (with several iterables):

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)]

Cartesian product (with one iterable and itself):

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)]
水波映月 2024-07-12 06:26:33

使用 itertools.permutations 标准库

import itertools
list(itertools.permutations([1, 2, 3]))

改编自此处演示了itertools.permutations 可能会被实现:

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:]

itertools.permutations 的文档中列出了几种替代方法。 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)

Use itertools.permutations from the standard library:

import itertools
list(itertools.permutations([1, 2, 3]))

Adapted from here is a demonstration of how itertools.permutations might be implemented:

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:]

A couple of alternative approaches are listed in the documentation of itertools.permutations. Here's one:

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

And another, based on 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)
一瞬间的火花 2024-07-12 06:26:33

我使用了基于阶乘数字系统的算法 - 对于长度为 n 的列表,您可以组合每个排列逐项进行,从每个阶段留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,因此可以使用阶乘系统中数字的数字作为索引。 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]]

此方法是非递归的,但在我的计算机上稍微慢一些,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)。 当我需要它时它就足够了,但它远不是 itertools.permutations 。

I used an algorithm based on the factorial number system- For a list of length n, you can assemble each permutation item by item, selecting from the items left at each stage. You have n choices for the first item, n-1 for the second, and only one for the last, so you can use the digits of a number in the factorial number system as the indices. This way the numbers 0 through n!-1 correspond to all possible permutations in lexicographic order.

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))

output:

[[0, 1, 2], [0, 2, 1], [1, 0, 2], [1, 2, 0], [2, 0, 1], [2, 1, 0]]

This method is non-recursive, but it is slightly slower on my computer and xrange raises an error when n! is too large to be converted to a C long integer (n=13 for me). It was enough when I needed it, but it's no itertools.permutations by a long shot.

喜爱纠缠 2024-07-12 06:26:33

在我看来,一个非常明显的方式也可能是:

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

A quite obvious way in my opinion might be also:

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
羁绊已千年 2024-07-12 06:26:33
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]
]
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

Output:

[
    [1, 2.0, 'three'], 
    [1, 'three', 2.0], 
    [2.0, 1, 'three'], 
    [2.0, 'three', 1], 
    ['three', 1, 2.0], 
    ['three', 2.0, 1]
]
挽清梦 2024-07-12 06:26:33

以下代码是给定列表的就地排列,作为生成器实现。 由于它只返回对列表的引用,因此不应在生成器之外修改列表。
该解决方案是非递归的,因此使用的内存较少。 也可以很好地处理输入列表中元素的多个副本。

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

The following code is an in-place permutation of a given list, implemented as a generator. Since it only returns references to the list, the list should not be modified outside the generator.
The solution is non-recursive, so uses low memory. Work well also with multiple copies of elements in the input list.

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
~没有更多了~
我们使用 Cookies 和其他技术来定制您的体验包括您的登录状态等。通过阅读我们的 隐私政策 了解更多相关信息。 单击 接受 或继续使用网站,即表示您同意使用 Cookies 和您的相关数据。
原文