You must log in or register to comment.
Reminds me of quantum-bogosort: randomize the list; check if it is sorted. If it is, you’re done; otherwise, destroy this universe.
Guaranteed to sort the list in nearly instantaneous time and with absolutely no downsides that are capable of objecting.
You still have to check that it’s sorted, which is O(n).
We’ll also assume that destroying the universe takes constant time.
In the universe where the list is sorted, it doesn’t actually matter how long the destruction takes!
It actually takes a few trillion years but its fine because we just stop considering the “failed” universes because they will be gone soon™ anyway.
Eh, trillion is a constant
Instead of destroying the universe, can we destroy prior, failed shuffle/check iterations to retain o(1)? Then we wouldn’t have to reload all of creation into RAM.
And the time complexity is only O(1)
I don’t think you can check if array of n elements is sorted in O(1), if you skip the check though and just assume it is sorted now (have faith), then the time would be constant, depending on how long you’re willing to wait until the miracle happens. As long as MTM (Mean Time to Miracle) is constant, the faithfull miracle sort has O(1) time complexity, even if MTM is infinite. Faithless miracle sort has at best the complexity of the algorithm that checks if the array is sorted.
Technically you can to down to O(0) if you assume all array are always sorted.
