wolffd@0: function [cov, covf] = gpcovar(net, x)
wolffd@0: %GPCOVAR Calculate the covariance for a Gaussian Process.
wolffd@0: %
wolffd@0: %	Description
wolffd@0: %
wolffd@0: %	COV = GPCOVAR(NET, X) takes  a Gaussian Process data structure NET
wolffd@0: %	together with a matrix X of input vectors, and computes the
wolffd@0: %	covariance matrix COV.  The inverse of this matrix is used when
wolffd@0: %	calculating the mean and variance of the predictions made by NET.
wolffd@0: %
wolffd@0: %	[COV, COVF] = GPCOVAR(NET, X) also generates the covariance matrix
wolffd@0: %	due to the covariance function specified by NET.COVARFN as calculated
wolffd@0: %	by GPCOVARF.
wolffd@0: %
wolffd@0: %	See also
wolffd@0: %	GP, GPPAK, GPUNPAK, GPCOVARP, GPCOVARF, GPFWD, GPERR, GPGRAD
wolffd@0: %
wolffd@0: 
wolffd@0: %	Copyright (c) Ian T Nabney (1996-2001)
wolffd@0: 
wolffd@0: % Check arguments for consistency
wolffd@0: errstring = consist(net, 'gp', 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: % Compute prior covariance
wolffd@0: if nargout >= 2
wolffd@0:   [covp, covf] = gpcovarp(net, x, x);
wolffd@0: else
wolffd@0:   covp = gpcovarp(net, x, x);
wolffd@0: end
wolffd@0: 
wolffd@0: % Add output noise variance
wolffd@0: cov = covp + (net.min_noise + exp(net.noise))*eye(ndata);
wolffd@0: