Example: abc
returns a,b,c,ab,ac,bc,ca,cb,abc,acb,bac,bca,cab,cba
Example: abc
returns a,b,c,ab,ac,bc,ca,cb,abc,acb,bac,bca,cab,cba
I think this question is quite interesting. As a Python fanatic, I cannot agree with @garry_qian’s answer. Since Python provides such a useful standard library, it would be a pity not to use it. From this standpoint, a concise, short (well, no...), but basically logically the same approach is as follows:
<code>>>> s = 'abc' >>> results = sorted([''.join(c) for l in range(len(s)) for c in permutations(s, l+1)]) ['a', 'b', 'c', 'ab', 'ac', 'ba', 'bc', 'ca', 'cb', 'abc', 'acb', 'bac', 'bca', 'cab', 'cba']</code>
There is nothing special about this writing method, it is just a for
list comprehension using two s.
At the same time, it returns a list of strings, which is closer to the original question.
But this sparked some curiosity in me, and I wanted to write it myself. At the same time, in other languages that do not have these permutation and combination tools, it may be easier to use the same logic to complete it.
The first thing I completed was the function about combination, which substitutes a string and returns all character combinations of all lengths, but does not arrange them:
<code>def get_combinations(string): combs = [] for i in range(1, 2**len(string)): pat = "{0:b}".format(i).zfill(len(string)) combs.append(''.join(c for c, b in zip(string, pat) if int(b))) return combs</code>
Test:
<code>>>> print get_combinations('abc') ['c', 'b', 'bc', 'a', 'ac', 'ab', 'abc']</code>
As expected, we got:
'c', 'b', 'a'
'bc', 'ac', 'ab'
length 3 'abc'
As expected, all combinations of various lengths are available.
This way of writing is definitely not the best, but I think the idea is interesting. The idea is to consider all combinations of 'abc'
. That means considering whether to take a
, whether to take b
and whether to take c
, so there are a total of 2*2*2 = 8
(2**len(string)
) combinations. , then doesn’t it correspond to:
<code>000 -> 都不取 001 -> 只取 c 010 -> 只取 b 011 -> 取 b c 100 -> 只取 a 101 -> 取 a c 110 -> 取 a b 111 -> 都取</code>
So in get_combinations
, I used a little trick to generate binary codes from 1 to 7, and then decided which characters should be used in each combination based on 0 and 1.
This has not yet completed the task. We still have to obtain the standard answer:
All permutations situations for each
combination
This produces get_permutations
the function:
<code>def get_permutations(clst): if len(clst)==1: return [clst[0]] results = [] for idx, c in enumerate(clst): results += [c+substr for substr in get_permutations(clst[:idx] + clst[idx+1:])] return results</code>
Test:
<code>>>> print get_permutations('abc') ['abc', 'acb', 'bac', 'bca', 'cab', 'cba']</code>
The logic is very simple, use recursive method to find all the permutations of 固定長度字元組合
.
With the above two functions, we can find the answer:
<code>>>> [perm for comb in get_combinations('abc') for perm in get_permutations(list(comb))] ['c', 'b', 'bc', 'cb', 'a', 'ac', 'ca', 'ab', 'ba', 'abc', 'acb', 'bac', 'bca', 'cab', 'cba']</code>
Conclusion:
Don’t reinvent tires, it will not only tire you out, but also make you look stupid
Life is short, I use Python
<code>import itertools chrs = 'abc' for i in range(len(chrs)): for combination in itertools.permutations(chrs, i + 1): print combination</code>
Since the php
and python
tags are marked at the same time, let’s write them in both ways. The logic is the same.
php
Code
<code class="php">function addChar($strs, $chars) { $result = []; foreach ($strs as $str) { foreach ($chars as $char) { $result[] = $str . $char; } } return $result; } $chars = ['a', 'b', 'c']; $group = []; $count = count($chars); for ($i = 1; $i <= $count; $i++) { if ($i == 1) { $group[$i] = addChar([''], $chars); } else { $group[$i] = addChar($group[$i - 1], $chars); } } // 合并数组 $result = call_user_func_array('array_merge', $group); var_dump($group);</code>
python
Code
<code class="python"># encoding:utf-8 def addChar(strs, chars): result = [] for str in strs: for char in chars: result.append(str + char) return result chars = ['a', 'b', 'c'] group = {} count = len(chars) for i in xrange(1, count + 1): if i == 1: group[i] = addChar([''], chars) else: group[i] = addChar(group[i - 1], chars) # 合并数组 result = [] for i in group: result += group[i] print result </code>
<code>result = [] def function(arg, string): global result if len(arg) >= len(string): return None for alphabet in string: if alphabet in arg: continue function(arg+alphabet, string) result.append(arg+alphabet) string = 'abc' for alphabet in string: result.append(alphabet) function(alphabet, string) print list(set(result))</code>
python2.7, the same as @garry_qian, I only found out after writing it. I am too lazy to look at other python solutions
<code class="python"># coding: utf-8 import itertools as t li = ['a', 'b', 'c'] tmp = [] for n in range(1, len(li) + 1): x = t.permutations(li, n) for i in x: tmp.append(''.join(i)) print tmp</code>
P(2,3)
P(3,3)
12 possibilities
Assume the length of the string is 2, then all the combinations are: 2! 2! / 1! = 4
Assume the length of the string is 3, then all the combinations are: 3! 3! / 1! 3! / 2! = 15
Assume the length of the string is 4, then all the combinations are: 4! 4! / 1! 4! / 2! 4! / 3! = 64
This formula can be carried out Promotion
n! n! / 1! n! / 2! ... n! / (n-1)!
The code will not be posted