🚀 go-pugleaf

G'day,
  I am working on a problem that requires fitting a simplex around a set of
data points in n-dimensions.  In other words, I have a set of
n-dimensional points and I'd like to find the smallest (least "volume") n-d
simplex (n+1 verices) that contains these points in its interior.  Any pointers
to previous work or ideas on how you have attacked similar problems would be
appreciated.  I am currently thinking of the problem in the light of linear
programming, where I'm trying to find the constraints (the bounding 
hyperplanes) given these interior points. For instance : 

In 2-d my simplex will be a triangle formed by three hyperplanes (lines).
I need to find the three lines that each separate the 2-d space into two
halfspaces (one with all my data and one with none), and together form the
triangle of minimum area that contain my data. And of course I want this to be
generalized to any dimension, be robust, easy to code, not memory intensive,
make me breakfast and wash my car :) 

Hope to hear from you soon.

Joe

-- 
-
JOSEPH WILLIS BOARDMAN
Email: joe@syd.deg.csiro.au
Phone: 61-2-887-8884