题目:
Given a collection of numbers, return all possible permutations.
For example,[1,2,3]
have the following permutations:[1,2,3]
, [1,3,2]
, [2,1,3]
, [2,3,1]
, [3,1,2]
, and [3,2,1]
.
Backtracking
链接: http://leetcode.com/problems/permutations/
题解:
求数组的全排列。需要用到DFS和Backtracking。 原理是从0到数组长度N,每次对之前加入的元素进行回溯。 注意此解法假定输入数组之中没有重复元素。
Time Complexity - O(n!), Space Complexity - O(n)。
public class Solution { public ArrayList> permute(int[] nums) { ArrayList > result = new ArrayList >(); if(nums == null || nums.length == 0) return result; Arrays.sort(nums); ArrayList list = new ArrayList (); helper(result, list, nums); return result; } private void helper(ArrayList > result, ArrayList list, int[] nums){ if(list.size() == nums.length){ //in this problem we assume no duplicate exists in input array result.add(new ArrayList (list)); return; } for(int i = 0; i < nums.length; i++){ if(list.contains(nums[i])) continue; list.add(nums[i]); helper(result, list, nums); list.remove(list.size() - 1); } } }
二刷:
Java:
一般来说可以分为两种做法,一种是DFS + backtracking, 另外一种是利用next permutation一样的做法来一个一个求出next permutation。
Next permutation做法:
为了使用Arrays.asList(new Nums),我们先把int[] nums转换为Integer[] newNums。然后新建一个method - boolean hasNextPermutation。把newNums排序以后,转换为ArrayList,加入到结果集中。接下来我们进入循环,当hasNext为true时,我们在hasNext的函数体里面已经完成了交换,把这时候的nums转换为ArrayList加入到结果集中,然后进行下一次判断。 最后返回结果。 速度比较慢。
Time Complexity - O(n!), Space Complexity - O(n)
public class Solution { public List> permute(int[] nums) { List
> res = new ArrayList<>(); if (nums == null || nums.length == 0) { return res; } Integer[] newNums = new Integer[nums.length]; for (int i = 0; i < nums.length; i++) { newNums[i] = nums[i]; } Arrays.sort(newNums); res.add(new ArrayList
(Arrays.asList(newNums))); while (hasNextPermutation(newNums)) { res.add(new ArrayList (Arrays.asList(newNums))); } return res; } private boolean hasNextPermutation(Integer[] newNums) { for (int i = newNums.length - 2; i >= 0; i--) { if (newNums[i] < newNums[i + 1]) { for (int j = newNums.length - 1; j > i; j--) { if (newNums[j] > newNums[i]) { swap(newNums, i, j); reverse(newNums, i + 1, newNums.length - 1); return true; } } } } return false; } private void swap(Integer[] nums, int i, int j) { int tmp = nums[i]; nums[i] = nums[j]; nums[j] = tmp; } private void reverse(Integer[] nums, int i, int j) { while (i < j) { swap(nums, i++, j--); } } }
DFS + backtracking(使用list.contains去重复):
我们也可以用DFS+Backtracking来做。这里需要建立一个helper函数permute和一个辅助List
Time Complexity - O(n!), Space Complexity O(n)。 (其实由于有这一步list.contains()递归复杂度应该是 (n * 1) *((n - 1) * 2) *((n - 2) * 3)...)应该等于 ∏(12*..。 *n2)
public class Solution { public List> permute(int[] nums) { List
> res = new ArrayList<>(); if (nums == null || nums.length == 0) { return res; } Arrays.sort(nums); List
onePerm = new ArrayList (); permute(res, onePerm, nums); return res; } public void permute(List > res, List
onePerm, int[] nums) { if (onePerm.size() == nums.length) { res.add(new ArrayList (onePerm)); return; } for (int i = 0; i < nums.length; i++) { if (onePerm.contains(nums[i])) { continue; } onePerm.add(nums[i]); permute(res, onePerm, nums); onePerm.remove(onePerm.size() - 1); } } }
DFS:
也可以用类似自底向上新建立List
- 我们先传入辅助函数一个空ArrayList<>()
- 遍历数组,当index i <= newPerm.size()时,我们把num[pos]这个元素分别加入到这个newPerm的从0 到 newPerm.size()的不同位置中去
- 然后再进行下一层dfs,继续加入nums中的下一个元素
- 这样做的坏处是空间复杂度会比较高,每一层的每次遍历都会新开一个ArrayList,假如nums很大的话,理论上有可能会触发更多次的garbage collection。但试验了几次以后比上面的两个solution要快,先存着当做一种不同的思路了。
- 还有一点是这里的code没有对nums排序,好像这种方法排序与不排序并没有什么关系....
Time Complexity - O(n!), Space Complexity - O(n!)
public class Solution { public List> permute(int[] nums) { List
> res = new ArrayList<>(); if (nums == null || nums.length == 0) { return res; } getPermutions(res, new ArrayList
(), nums, 0); return res; } public void getPermutions(List > res, List
onePerm, int[] nums, int pos) { if (onePerm.size() == nums.length) { res.add(new ArrayList (onePerm)); return; } for (int i = 0; i <= onePerm.size(); i++) { List newPerm = new ArrayList<>(onePerm); newPerm.add(i, nums[pos]); getPermutions(res, newPerm, nums, pos + 1); } } }
三刷:
还是使用了next permutation的方法。需要再练习练习。还有一些方法,比如不断地把第i位的元素插入到 i - 1的结果中去。 或者不断递归swap start和i并且回溯,都可以完成。Discuss区的大神们写得更好。
查了一下资料,更好的方法应该是Johnson-Trotter法,根据上一个permutation的奇偶性来升序或者降序地添加新元素,也是O(n!)的时间复杂度。但这么做可能导致结果是无序的。
Java:
使用Next Permutation:
public class Solution { public List> permute(int[] nums) { List
> res = new ArrayList<>(); if (nums == null || nums.length == 0) return res; Arrays.sort(nums); do { List
permu = new ArrayList<>(); for (int num : nums) permu.add(num); res.add(permu); } while (hasNextPermutation(nums)); return res; } private boolean hasNextPermutation(int[] nums) { int len = nums.length; for (int i = len - 2; i >= 0; i--) { if (nums[i] < nums[i + 1]) { for (int j = len - 1; j > i; j--) { if (nums[j] > nums[i]) { swap(nums, i, j); reverse(nums, i + 1, len - 1); return true; } } } } return false; } private void swap(int[] nums, int i, int j) { int tmp = nums[i]; nums[i] = nums[j]; nums[j] = tmp; } private void reverse(int[] nums, int i, int j) { while (i < j) swap(nums, i++, j--); } }
通过不断swap: 速度非常快
public class Solution { public List> permute(int[] nums) { List
> res = new ArrayList<>(); permute(res, nums, 0); return res; } private void permute(List
> res, int[] nums, int pos) { if (pos == nums.length) { List
list = new ArrayList<>(); for (int i = 0; i < nums.length; i++) list.add(nums[i]); res.add(list); return; } for (int i = pos; i < nums.length; i++) { swap(nums, pos, i); permute(res, nums, pos + 1); swap(nums, pos, i); } } private void swap(int[] nums, int i, int j) { int tmp = nums[i]; nums[i] = nums[j]; nums[j] = tmp; } }
通过不断insert下一个元素
public class Solution { public List> permute(int[] nums) { List
> res = new ArrayList<>(); if (nums.length == 0) return res; res.add(new ArrayList<>()); for (int i = 0; i < nums.length; i++) { List
> newRes = new ArrayList<>(); for (int j = 0; j <= i; j++) { for (List
l : res) { List list = new ArrayList<>(l); list.add(j, nums[i]); newRes.add(list); } } res = newRes; } return res; } }
Reference:
http://introcs.cs.princeton.edu/java/23recursion/Permutations.java.html
https://en.wikipedia.org/wiki/Heap%27s_algorithm
https://leetcode.com/discuss/62270/simple-python-solution-68ms
https://leetcode.com/discuss/55127/java-solution-easy-to-understand-backtracking
https://leetcode.com/discuss/55418/java-clean-code-two-recursive-solutions
https://leetcode.com/discuss/19510/my-ac-simple-iterative-java-python-solution
https://leetcode.com/discuss/29483/share-my-short-iterative-java-solution
https://leetcode.com/discuss/18212/my-elegant-recursive-c-solution-with-inline-explanation
http://stackoverflow.com/questions/5363619/complexity-of-recursive-string-permutation-function
https://en.m.wikipedia.org/wiki/Steinhaus%E2%80%93Johnson%E2%80%93Trotter_algorithm
http://introcs.cs.princeton.edu/java/23recursion/JohnsonTrotter.java.html