2 回答
TA贡献1797条经验 获得超4个赞
关闭迭代模拟:
def permutations(A):
P = []
permutationsI(A, P)
print(P)
def permutationsI(A, perms):
stack = [(A, [])]
while len(stack):
first, last = stack.pop()
if len(first):
for i in range(len(first)):
stack.append((first[:i] + first[i+1:],last + [first[i]]))
else:
perms.append(last)
permutations([1,2,3])
>>[[3, 2, 1], [3, 1, 2], [2, 3, 1], [2, 1, 3], [1, 3, 2], [1, 2, 3]]
TA贡献1821条经验 获得超4个赞
一个完全递归的函数应该是:
def permutations_comp_recursive(first, last, perms, i):
if len(first) == 0:
perms.append(last)
elif i == len(first):
pass
else:
permutations_comp_recursive(first, last, perms, i+1)
if first:
permutations_comp_recursive(
first[:i]+first[i+1:],
last + [first[i]],
perms, 0)
为了获得良好的性能,我推荐numpy 解决方案。
编辑 1:现在以下应该是尾递归的,使用列表理解。这使用了python 中尾递归的解决方法(省略了最后 2 个参数 - 结果作为返回值传递):
import itertools as it
class Recurse(Exception):
def __init__(self, *args, **kwargs):
self.args = args
self.kwargs = kwargs
def recurse(*args, **kwargs):
raise Recurse(*args, **kwargs)
def tail_recursive(f):
def decorated(*args, **kwargs):
while True:
try:
return f(*args, **kwargs)
except Recurse as r:
args = r.args
kwargs = r.kwargs
continue
return decorated
@tail_recursive
def permutations_tail_recursive(first, last, direct=False):
if len(first) == 0 or not all(first):
return last
else:
l = len(first[0]) if direct else len(first)
if last:
return recurse([fi[:i]+fi[i+1:] for fi, i in it.product(first, range(l))],
[last[j] + first[j][i] for j, i in it.product(range(len(last)), range(l))], True)
else:
return recurse([first[:i]+first[i+1:] for i in range(l)],
[first[i] for i in range(l)], True)
这未优化并使用循环。我不确定这和上面没有循环的代码是否可以组合 - 可能会再次研究它。itertools.permutations 可用于此应用程序。
添加回答
举报