in reply to Re^10: Algorithm RFC: fast (pseudo-)random shuffle with no repetition
in thread Algorithm RFC: fast (pseudo-)random shuffle with no repetition

There is nothing I have to be convinced of. It was clear that with this method at some point end is. With ((1)x 9, (2) x 8) the red area begins on the relatively weak machine on which I tried this via IPad/SSH: 0m0,003s in the best and 0m41,493s in the worst case so far. With ((1)x 51, (2) x 50) that's completely hopeless there. With the OP's example (15), the result was actually quite acceptable - unlike what he expected predicted by him.

«The Crux of the Biscuit is the Apostrophe»

  • Comment on Re^11: Algorithm RFC: fast (pseudo-)random shuffle with no repetition