camir-aes2014: toolboxes/FullBNT-1.0.7/netlab3.3/dist2.m annotate

first hg version after svn

rev	line source
wolffd@0	1 function n2 = dist2(x, c)
wolffd@0	2 %DIST2 Calculates squared distance between two sets of points.
wolffd@0	3 %
wolffd@0	4 % Description
wolffd@0	5 % D = DIST2(X, C) takes two matrices of vectors and calculates the
wolffd@0	6 % squared Euclidean distance between them. Both matrices must be of
wolffd@0	7 % the same column dimension. If X has M rows and N columns, and C has
wolffd@0	8 % L rows and N columns, then the result has M rows and L columns. The
wolffd@0	9 % I, Jth entry is the squared distance from the Ith row of X to the
wolffd@0	10 % Jth row of C.
wolffd@0	11 %
wolffd@0	12 % See also
wolffd@0	13 % GMMACTIV, KMEANS, RBFFWD
wolffd@0	14 %
wolffd@0	15
wolffd@0	16 % Copyright (c) Ian T Nabney (1996-2001)
wolffd@0	17
wolffd@0	18 [ndata, dimx] = size(x);
wolffd@0	19 [ncentres, dimc] = size(c);
wolffd@0	20 if dimx ~= dimc
wolffd@0	21 error('Data dimension does not match dimension of centres')
wolffd@0	22 end
wolffd@0	23
wolffd@0	24 n2 = (ones(ncentres, 1) * sum((x.^2)', 1))' + ...
wolffd@0	25 ones(ndata, 1) * sum((c.^2)',1) - ...
wolffd@0	26 2.(x(c'));
wolffd@0	27
wolffd@0	28 % Rounding errors occasionally cause negative entries in n2
wolffd@0	29 if any(any(n2<0))
wolffd@0	30 n2(n2<0) = 0;
wolffd@0	31 end

Mercurial > hg > camir-aes2014