Cyclic sort

What if I wanted to sort a list of integers, but with the added condition that the list was cyclical and the sorted data could "start" at any point? e.g. the list [7, 4, 2, 5, 3] could be sorted as [2, 3, 4, 5, 7], [5, 7, 2, 3, 4], [4, 5, 7, 2, 3], etc.

Would an algorithm that implements this sort be more efficient than a standard sort? It seems like since there are now n solutions instead of just 1, sorting the list should be n times easier and thus only take O(log n) time. But I think that's probably bullshit.

(define (cyclic-sort lis)
  (apply circular-list (sort lis)))

sub oh_en_sort {
    for (@_) { @a[$_] = $_; }
    return @a;
}

boom, O(n) sort. For 1..n sequences only, though. You could add a grep{/./} in there to remove empty array elements but that would increase the complexity.

_{Fibonacci buttsort.}

/prog/ already invented sleepsort, the only sort you'll ever need.

>>3
That is radix sort. It has kN complexity, not O(N)
http://en.wikipedia.org/wiki/Radix_sort

>>3

What the fuck language is that?  Ruby?  Perl?

>>7
That's perl, baby. @_ is the implicit array of arguments provided to the most local sub or function. It's basically TECO, only somehow uglier.

it's the same "standard" (dunno what you mean as 'standard' though but w/e) sort just with some additional parameter where to start

>seems like since there are now n solutions instead of just 1

and only 1 solution is correct for any given parameter


a = [7, 4, 2, 5, 3]
b, c, n = [], [], 3
for i in a:
    if i > n:
        b += [i]
    else:
        c += [i]
b.sort()
c.sort()
a = b + c

>>6
What the fuck is $kN$ and how is it different from $n$?

 Ruitomo? Killing curse that's not unique to you is good enough.

How are little girls disgusting? If that's not it, how is a grown man masturbating to little girls any more disgusting than another grown man masturbating to women?