wolffd@0: function [post, a] = gmmpost(mix, x)
wolffd@0: %GMMPOST Computes the class posterior probabilities of a Gaussian mixture model.
wolffd@0: %
wolffd@0: %	Description
wolffd@0: %	This function computes the posteriors POST (i.e. the probability of
wolffd@0: %	each component conditioned on the data P(J|X)) for a Gaussian mixture
wolffd@0: %	model.   The data structure MIX defines the mixture model, while the
wolffd@0: %	matrix X contains the data vectors.  Each row of X represents a
wolffd@0: %	single vector.
wolffd@0: %
wolffd@0: %	See also
wolffd@0: %	GMM, GMMACTIV, GMMPROB
wolffd@0: %
wolffd@0: 
wolffd@0: %	Copyright (c) Ian T Nabney (1996-2001)
wolffd@0: 
wolffd@0: % Check that inputs are consistent
wolffd@0: errstring = consist(mix, 'gmm', x);
wolffd@0: if ~isempty(errstring)
wolffd@0:   error(errstring);
wolffd@0: end
wolffd@0: 
wolffd@0: ndata = size(x, 1);
wolffd@0: 
wolffd@0: a = gmmactiv(mix, x);
wolffd@0: 
wolffd@0: post = (ones(ndata, 1)*mix.priors).*a;
wolffd@0: s = sum(post, 2);
wolffd@0: if any(s==0)
wolffd@0:    warning('Some zero posterior probabilities')
wolffd@0:    % Set any zeros to one before dividing
wolffd@0:    zero_rows = find(s==0);
wolffd@0:    s = s + (s==0);
wolffd@0:    post(zero_rows, :) = 1/mix.ncentres;
wolffd@0: end
wolffd@0: post = post./(s*ones(1, mix.ncentres));