Path: utzoo!mnetor!uunet!husc6!mailrus!ames!eos!jbm
From: jbm@eos.UUCP (Jeffrey Mulligan)
Newsgroups: comp.graphics
Subject: Re: Algorithm wanted: Circle enclosing points
Message-ID: <498@eos.UUCP>
Date: 5 Apr 88 19:31:50 GMT
References: <494@eos.UUCP>
Organization: NASA Ames Research Center, California
Lines: 32

Thanks to Guy Jacobson for posting a complete reference to
this apparently well-studied problem.  In light of this,
I don't know whether there is much point in correcting
my previous erroneous solution,

From article <494@eos.UUCP>, by jbm@eos.UUCP (Jeffrey Mulligan):
> 1)  Make a list of all triples of points (there are N choose 3 of them).
> 
> 2)  Treating each triple of points as the vertices of a triangle,

2a)  If the triangle is right or acute:

> 	compute the radius of the circumscribing circle.

2b) if the triangle is obtuse, the circle's diameter is the side
opposite the obtuse angle.

> 
> 3)  The largest circumcircle is the desired result.
> 

s/circumcircle/circle



Guess I really put my foot in it after that posting about naive answers...

-- 

	Jeff Mulligan (jbm@ames-aurora.arpa)
	NASA/Ames Research Ctr., Mail Stop 239-3, Moffet Field CA, 94035
	(415) 694-5150