home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.math
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!howland.reston.ans.net!spool.mu.edu!yale.edu!ira.uka.de!math.fu-berlin.de!news.tu-chemnitz.de!mb3.tu-chemnitz.de!mapool14.mathematik.tu-chemnitz.de!muelleru
- From: muelleru@Mathematik.TU-Chemnitz.DE (Ulrich Mueller)
- Subject: Re: proof wanted 2
- Message-ID: <muelleru.17@Mathematik.TU-Chemnitz.DE>
- Sender: inews@mb3.tu-chemnitz.de (Internet news)
- Organization: University of Technology Chemnitz, FRG
- References: <1993Jan8.195646.1694@cc.umontreal.ca> <1ikq9eINNmue@roundup.crhc.uiuc.edu> <1993Jan9.193759.3671@Princeton.EDU> <1iorntINNoal@skeena.ucs.ubc.ca>
- Date: Mon, 11 Jan 1993 14:23:28 GMT
- Lines: 17
-
- In article <1iorntINNoal@skeena.ucs.ubc.ca> liuli@unixg.ubc.ca (Li Liu) writes:
-
- >The claim that there is a closest point in C (a close set) to a point
- >x (a point outside C) is not true for general metric spaces.
-
- >A simple example is to take the metric space to be { 1/n | n is positive
- >integers }, under the usual metric d(a,b)= |a-b|. Let C be the space
- >itself. C is closed. Let x be 0. There is no point in C that is closet to
- >zero.
-
- No, I think that example isn't true.
- Because 0 isn't a point of your metric space. If it was one, so {1/n}
- was not closed.
-
- I think, in every closed non-empty subset C of a metric space M there is a
- point with the minimal distance to a point x from M-C.
- Prove I just sent.
-