home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!dtix!uranus!tecsun1!zogwarg!hoey
- From: hoey@zogwarg.etl.army.mil (Dan Hoey)
- Newsgroups: comp.programming
- Subject: Re: Locating duplicates
- Message-ID: <1293@zogwarg.etl.army.mil>
- Date: 7 Jan 93 17:01:06 GMT
- References: <1993Jan5.043456.29130@micor.ocunix.on.ca> <1993Jan5.224141.21639@organpipe.uug.arizona.edu> <1993Jan7.021146.2290@cs.cornell.edu>
- Organization: Naval Research Laboratory, Washington, DC
- Lines: 15
-
- karr@cs.cornell.edu (David Karr) writes:
- >A shortcoming of course is that the running-time analysis is valid
- >only if you can repeatedly allocate large amounts of zeroed-out memory
- >in a small (essentially constant) time.
-
- You can actually skip the zeroing out at a cost of a constant factor
- at access time. It's well-known; email me if you're interested.
-
- There's a known linear-time solution for the element uniqueness
- problem on the real-ram-with-floor model (none of this bogus 32-bit
- integer restriction) that uses a kind of hash table in this way. It
- was in JACM in the seventies, if I recall.
-
- Dan Hoey
- Hoey@AIC.NRL.Navy.Mil
-