Quote:
Originally Posted by PeetSoft
Google wasn't my friend today ;(

I feel your pain.
There are a whole bunch of scientific papers on what is known as
the "Euclidean distance matrices completion problem" most of them including the phrase "NP Hard".
I've read that a generalized solution (or set of solutions) to this problems would have implications for
molecular conformation in bioinformatics, dimensionality reduction in machine learning and statistics, and even the problem of wireless sensor network localization.
Hopefully you've checked around the MatLab forums.
Here's
one post that says:
Quote:
hi anyone know what is the equation for euclidean distance between two matrices..
One thing you can do is to reshape the matrices into
vectors and then apply the usual Euclidean distance measure.
It doesn't really matter how you do the reshaping, but if
you wish to create vectors from a matrix going row by row then
do this (for two matrices A and B):
A2 = reshape(A',prod(size(A)),1); % makes column vectors
B2 = reshape(B',prod(size(B)),1);
dist = sqrt(dot(A2B2,A2B2));

Thinking about ways which might help your googling:
Euclidean (utilizing root sumofsquares of differences) is only type of distance calculation.
Often this method is found in science/math/coding documents which contain other methods like: "manhattan" "canberra" "minkowski"
Hopefully these keywords can help with your searching..