home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!munnari.oz.au!bunyip.cc.uq.oz.au!uqcspe!cs.uq.oz.au!warwick
- From: warwick@cs.uq.oz.au (Warwick Allison)
- Newsgroups: rec.games.programmer
- Subject: Re: Advice needed for fastest sort algorthm
- Message-ID: <11056@uqcspe.cs.uq.oz.au>
- Date: 16 Nov 92 02:37:32 GMT
- References: <6759@news.duke.edu> <BxD5pt.DyI@news.cso.uiuc.edu> <BxG0MC.7yK@chinet.chi.il.us> <10991@uqcspe.cs.uq.oz.au> <1drd06INNnji@calvin.usc.edu>
- Sender: news@cs.uq.oz.au
- Reply-To: warwick@cs.uq.oz.au
- Lines: 29
-
- dank@calvin.usc.edu (Dan King) writes:
-
- >Quoting from The Art of Computer Programming, Volume 3 (pp176-178):
- >"[r]adix sorting is usually more efficient than any other known
- >sorting method for internal sorting."
-
- Is that the first edition (1968), or second edition (1973).
- I hope the references haven't dated too much :-).
-
- Seriously, it's these exact statement that I'm arguing against. Saying that ANY
- algorithm for sorting is the best, or "usually more efficient than any other" is
- simply crazy. Each algorithm has strengths and weaknesses. Radix sort isn't
- even APPLICABLE in most cases.
-
-
- >Speaking about his example program, Knuth says,
- >"A p-pass version of the program would take (11p-1)N+O(pM) units of
- >time." [p is the number of keys and M is the number of radix piles]
-
- ie. if p is too large, forget it. too large = O(N)... or even O(sqrt(N)).
- (where N is the EXPECTED number of items for sorting)
-
- --
- Warwick
- --
- _-_|\ warwick@cs.uq.oz.au /Disclaimer:
- / * <-- Computer Science Department, /
- \_.-._/ University of Queensland, / C references are NULL && void*
- v Brisbane, Australia. /
-