Re: Fortran to find nearest point from set in 3-D
- From: David.Paterson@xxxxxxxx
- Date: 16 Apr 2006 19:18:29 -0700
octree I know about, but k-d tree etc. are new to me. By the way, I
will have to do this 'proximal-point' problem hundreds of times with
different sets of points B.
MOST of B can be eliminated at a glance. ... "At a glance" requires more computational effort than glancing
I figure that, too.
presumably you're already computed the distances from the grid cubes tothe old point ... discrete approximation to a convex set
Huh?
Do you know anything more about the sets?
Yes. Set A (the small one) comes from the triangular discretisation on
a set of surfaces and Set B (the big one) comes from the the corners of
a tetrahedral discretisation of 3-D space. Unfortunately all the
connectivity has been lost in the form that I have access to.
This means that the points in B are roughly equidistance apart, I may
even be able to give an 'a priori' maximum distance apart. If it helps,
call it 'D' for distance. I can also put hard upper and lower bounds on
each of the x, y and z of the triples (x,y,z) in B.
And if it's good enough to do this in the sup (max) norm instead of theL^2 norm
L^2 won't take that much more time provided I remember not to take
square roots :-)
.
- Follow-Ups:
- Re: Fortran to find nearest point from set in 3-D
- From: Rob McDonald
- Re: Fortran to find nearest point from set in 3-D
- References:
- Fortran to find nearest point from set in 3-D
- From: David . Paterson
- Re: Fortran to find nearest point from set in 3-D
- From: Ronald Bruck
- Fortran to find nearest point from set in 3-D
- Prev by Date: Unsigned arithmetic
- Next by Date: Re: Fortran to find nearest point from set in 3-D
- Previous by thread: Re: Fortran to find nearest point from set in 3-D
- Next by thread: Re: Fortran to find nearest point from set in 3-D
- Index(es):