Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.lang.forth > #23951
| Newsgroups | comp.lang.forth |
|---|---|
| Date | 2013-06-26 11:59 -0700 |
| References | <2013Jun26.181725@mips.complang.tuwien.ac.at> |
| Message-ID | <b232eaf4-f3c7-4bc7-afb2-54eedbbd95a3@googlegroups.com> (permalink) |
| Subject | Re: som probabilistic sorts |
| From | mhx@iae.nl |
anton@mips.complang.tuwien.ac.at (Anton Ertl) writes Re: som probabilistic sorts > Inspired by the recent flurry of sorting algorithms, I devised some of > my own, all probabilistic and pretty simple, but still a surprising > amount of of code. All code plus some scaffolding at > <http://www.complang.tuwien.ac.at/forth/programs/badsort.fs> Bad-sort is simply unbearable (even for 10 elements). All array elements random, less than 100. Mediocre-sort speeds up nicely. Combining maybe-exchange with simple-sort doesn't give a consistent improvement over simple-sort alone. FORTH> bm *** Sorts with /size = 1000 *** bubble-sort : 0.005 seconds elapsed. simple-sort : 0.003 seconds elapsed. so-so-sort : 0.700 seconds elapsed. mediocre-sort : 0.035 seconds elapsed. ok FORTH> bm *** Sorts with /size = 4000 *** bubble-sort : 0.054 seconds elapsed. simple-sort : 0.020 seconds elapsed. so-so-sort : 155.069 seconds elapsed. mediocre-sort : 0.392 seconds elapsed. ok -marcel
Back to comp.lang.forth | Previous | Next — Previous in thread | Find similar
som probabilistic sorts anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2013-06-26 16:17 +0000 Re: som probabilistic sorts mhx@iae.nl - 2013-06-26 11:59 -0700
csiph-web