Re: SQL challenge

From: Mikito Harakiri <mikharakiri_at_iahu.com>
Date: Tue, 13 Jul 2004 11:48:21 -0700
Message-ID: <RjWIc.37$KC5.155_at_news.oracle.com>

"Jan Hidders" <jan.hidders_at_REMOVETHIS.pandora.be> wrote in message news:pan.2004.07.13.17.45.22.985796_at_REMOVETHIS.pandora.be...
> Well, the problem was to find the *minimal* set. (Actually I didn't
> define that properly, what I meant was the set with the least number of
> edges.) If I give you the corners of a cube plus some extra points on its
> surface and its edges then I still just want the edges between the corner
> points of the cube. If you simply take the edges of your solution then
> there are many more edges. For example, if there is point in the middle of
> a plane of the convex hull then you will have (redundant) edges from and
> to this point. Or if there is a node in the middle of an edge of the
> convex hull then you will have (redundant) edges from and to this point.
> These need to be removed.
>
> You're not there yet. :-)

But this is just planar version of the same conveax hull problem!

Another glitch is for each triangular face (p1,p2,p3) the output would have them in all permutations

 (p1,p2,p3)
 (p2,p1,p3)
 (p1,p3,p1)

...
so that I have to take care of this too.

Solving these small problems and integrating them into the answer doesn't look particularly challenging to me:-) but, i'm afraid, it might lengthen the final solution a bit.

Is there an elegant SQL statement? If you say yes, then, I would think hard finding it:-)

BTW, after posting the first attempt, I gooled Jims 2002 paper. He seems to consider planar case only, right? Received on Tue Jul 13 2004 - 20:48:21 CEST