Finding Permutations�Easier and Faster
Introduction
This article explains the technique of finding permutations in a simple and fast manner. It also provides the source code for the same.
Explanation
For a given string of length N, there are actually N! Permutations. The technique we apply here is to find the unique of those permutations and display them.
Let's take this string: 123. If we rotate this string in a circular manner, we get 231 and 312. The point here is, if we find 1 string of the N! Permutations, we can produce N permutations by rotating it circularly. This reduces the time, which is proportional to N. The greater the value of N, the more the algorithm is optimized.
In short, the algorithm finds a number and rotates the number circularly N times to get N numbers.
Lets apply the algorithm. Lets take string {12} of length 2 (N=2). The unique pattern is 12 and if we rotate we get 21.
Let's take the string {123} of length 3 (N=3). The unique patterns are as follows:
1(23): 123 is unique pattern resulting in (123)
1(32): Rotate the pattern (23) u get another pattern (32) resulting in a pattern (132)
(231): By rotating the pattern (123)
(312): By rotating the pattern (123)
(321): By rotating the pattern (132)
(213): By rotating the pattern (132)
The output patterns are:
| Level 1 | Level 2 | Permutations |
| 123 | ||
| 1(23) | 123 | |
| 231 | ||
| 312 | ||
| 1(32) | 132 | |
| 321 | ||
| 213 |
If we apply the same algorithm for the string {1234} of length 4 (N=4), the patterns are as below:
| Level 1 | Level 2 | Level 3 | Permutations |
| 1234 | |||
| 12(34) | |||
| 1(234) | 1234 | ||
| 2341 | |||
| 3412 | |||
| 4123 | |||
| 1(342) | 1342 | ||
| 3421 | |||
| 4213 | |||
| 2134 | |||
| 1(423) | 1423 | ||
| 4231 | |||
| 2314 | |||
| 3142 | |||
| 12(43) | |||
| 1(243) | 1243 | ||
| 2431 | |||
| 4312 | |||
| 3124 | |||
| 1(432) | 1432 | ||
| 4321 | |||
| 3214 | |||
| 2143 | |||
| 1(324) | 1324 | ||
| 3241 | |||
| 2413 | |||
| 4132 |
Comparison
The traditional algorithm goes and finds all the permutations where, as in this algorithm, we find only the unique permutation and produce the other permutations by rotating the original permutation in a circular manner.
Conclusion
The algorithm finds only the unique numbers and finds the other (N) permutations by rotating the original permutation circularly. Hence, it increases the performance.

Comments
Impressed
Posted by Shivinder Singh Narr on 03/12/2013 11:59amHey, I have implemented your algorithm of permutation. It creates all permutations of a string. It was a great idea. If you observe, it's time complexity is very less than traditional methods. I am saying this because I have used inbuilt permutation function in python. I analysed the time and observed that This algorithm implemented in C, requires less time than other methods. According to my program, it's time complexity is O(n^4). If you have anything to share about much more faster algorithm than this, please reply to my email address.
Replybacklinks
Posted by backlinks on 09/05/2012 11:17pmWrite more, thats all I have to say. Literally, it seems as though you relied on the video to make your point. You obviously know what youre talking about, why throw away your intelligence on just posting videos to your blog when you could be giving us something informative to read?
Replytry
Posted by try on 07/13/2012 01:41pmtry...
Replysolution
Posted by naveen kumar on 07/13/2012 01:39pmEx : 1234 rotate once right ... 4123,3412,2341,1234. reverse these numbers.. 3214,2143,1432,4321. Fix first digit of original number and rotate rest to right we get 1423... apply same algo. here. 3142,2314,4231,1423. reverse these number... 2413,4132,1324,3241. Rotate last time 1342 (Total rotations length-1 times).. 2134,4213,3421,1342.. reverse.. 4312,3124,1243,2431. Final Answere 24 values : 4123,3412,2341,1234. 3214,2143,1432,4321. 3142,2314,4231,1423. 2413,4132,1324,3241. 2134,4213,3421,1342. 4312,3124,1243,2431.
Replygood algo
Posted by mail_suka@rediffmail.com on 07/15/2004 11:24amhi, nice algo. btw, here is a interesting prob for u. how many ways, we can arrange 1 to n (n can be 10, 100 or 1000, any number) so that no number comes in its original position? (meaning 2 shud not appear in 2nd position, after arrangement). consider space, time comlpexity when u design the algo. regards;
ReplyGood one
Posted by kandukondein on 07/10/2004 02:28pmHey Thats a good one. A good tutorial for beginners. Regards;
Reply