One more, maybe: http://rosettacode.org/wiki/Sutherland-Hodgman_polygon_clipping#J
(The 'isinside' and 'intersection' verbs in the inner loop are probably too complicated. But, perhaps if they were unfolded and rewritten to be not so "j-like"?) -- Raul On Thu, May 26, 2016 at 1:59 PM, Henry Rich <[email protected]> wrote: > I did look at the other ones you originally mentioned after my 'top 3' guess > blew up. > > It's hard to work with programs that take a very short time to run, because > I'm not sure what is being timed. And some of them give wildly varying > times, even on constant input - odd. > > Of the programs given so far, I can use > > http://rosettacode.org/wiki/Bitmap/Midpoint_circle_algorithm#J (30% > improvement with new code) > http://rosettacode.org/wiki/Knuth%27s_power_tree#J (18% improvement) > http://rosettacode.org/wiki/Levenshtein_distance#J(24% on 1000-byte strings) > > That might be enough to call a test suite, but if you find others I'll check > them too. > > Henry Rich > > > > > > > > > On 5/26/2016 8:19 AM, Raul Miller wrote: >> >> With only one example, I'm going to have to do a bit more guessing >> than I'd like, but ok. If this was easy, it would be done already... >> >> You might also try running the maze solver with a larger maze. >> >> And, I guess I should also ask if it's specifically the maze *solver* >> which you find useful or the generator (the part also presented at >> http://rosettacode.org/wiki/Maze_generation#J ) or both which you find >> useful? >> >> Anyways, here's some things which might be "close enough": >> >> http://rosettacode.org/wiki/Bitmap/Midpoint_circle_algorithm#J >> http://rosettacode.org/wiki/Hofstadter_Figure-Figure_sequences#J >> http://rosettacode.org/wiki/K-d_tree#J >> http://rosettacode.org/wiki/Knuth%27s_power_tree#J >> http://rosettacode.org/wiki/Runge-Kutta_method#J >> http://rosettacode.org/wiki/Stable_marriage_problem#J >> http://rosettacode.org/wiki/Voronoi_diagram/J/Delaunay_triangulation >> (the convex hull part). >> >> Also... since two of the initial guesses at the "top 3" were actually >> not suitable, does that mean that some of the other, rejected, >> candidates might have been suitable? For example, it seems to as if >> that levenshtein algorithm is very "fortran-like". >> >> I am also wondering if the trial division primality test at >> http://code.jsoftware.com/wiki/Essays/Primality_Tests has any >> relevance? Or are arbitrary precision integers so slow that they mask >> any gains here? >> >> Thanks, >> > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
