http://www.perlmonks.org?node_id=554502


in reply to Re^3: possible combinations in sequence
in thread possible combinations in sequence

Just when you thought it couldn't get any faster...
Here's a comparison of the revised roboticus version (with ikegami's speed suggestion):
source: horse:cow:dog:cat Rate ruzam roboticus roboticus2 ruzam 4695/s -- -80% -86% roboticus 23049/s 391% -- -33% roboticus2 34607/s 637% 50% -- source: horse Rate ruzam roboticus2 roboticus ruzam 64954/s -- -71% -76% roboticus2 223392/s 244% -- -16% roboticus 266354/s 310% 19% -- source: horse:cat Rate ruzam roboticus roboticus2 ruzam 26625/s -- -72% -76% roboticus 96711/s 263% -- -13% roboticus2 111706/s 320% 16% -- source: horse:cow:cat Rate ruzam roboticus roboticus2 ruzam 11023/s -- -77% -82% roboticus 46987/s 326% -- -24% roboticus2 62225/s 465% 32% -- source: horse:cow:dog:cat:mouse Rate ruzam roboticus roboticus2 ruzam 1913/s -- -82% -89% roboticus 10826/s 466% -- -39% roboticus2 17652/s 823% 63% -- source: horse:cow:dog:cat:rat:mouse Rate ruzam roboticus roboticus2 ruzam 798/s -- -85% -91% roboticus 5253/s 559% -- -43% roboticus2 9221/s 1056% 76% --

The new function is now an magnitude faster than my original attempt. That's no small change. Is this the limit of optimizing or can yet more juice be squeezed out of this function?

Replies are listed 'Best First'.
Re^5: possible combinations in sequence
by roboticus (Chancellor) on Jun 13, 2006 at 12:23 UTC
    Well ... I've tried a few things off and on for a couple of days. I can't make it one whit faster.

    I've found a good few ways to make it slower, though! 8^)

    --roboticus