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»
In reply to Re^11: Algorithm RFC: fast (pseudo-)random shuffle with no repetition
by karlgoethebier
in thread Algorithm RFC: fast (pseudo-)random shuffle with no repetition
by Anonymous Monk
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |