[EMAIL PROTECTED] wrote: > Steve Wampler <[EMAIL PROTECTED]> skribis: >> That's essentially a Goedel numbering scheme. It's clever. It could >> generate >> large numbers if the input has long phrases (especially with lots of letters >> near the end of the alphabet!) I don't know if there's a performance >> advantage >> over using the letters themselves to form the signature or not. > > The strings are equivalent to base-256 so you could easily convert > them to integers—especially if Icon had EQUIVALENCE statements!
Right - but why? A string signature also uniquely identifies the equivalence classes - there's no need to convert it an integer that ends up being used the same way. I'm tempted to try and measure the performance of both approaches by writing two versions of the getClassSignature() method - one with Goedel numbering and the other using the 'characters in lexical order' to see if there's a performance benefit either way, but my instinct is that (at least for this puzzle) either solution is so fast as to be indistinguishable in performance. So I'd have to find a puzzle (maybe feed the program /usr/dict/web2 instead of the names of states) where we could see the difference... -- Steve Wampler -- [EMAIL PROTECTED] The gods that smiled on your birth are now laughing out loud. ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Unicon-group mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/unicon-group
