function [R] = spm_vb_get_R(slice,h0)
% Get posterior correlation matrix for regression coefficients
% FORMAT [R] = spm_vb_get_R(slice,h0)
% 
% slice  - data structure (see spm_vb_glmar)
% 
% R      - posterior correlation matrix of regression coefficients
%__________________________________________________________________________
% Copyright (C) 2005-2014 Wellcome Trust Centre for Neuroimaging

% Will Penny 
% $Id: spm_vb_get_R.m 6079 2014-06-30 18:25:37Z spm $


lambda = h0(end);
k = slice.k;
p = slice.p;
T = slice.T;
X = slice.X;

% This is described in the section 'approximating posterior
% covariance matrices' in paper VB3
if slice.p==0
    A = X'*X;
else
    a = h0(1:slice.p);
    x_err         = zeros(T-p,k);
    XtilVXtil     = zeros(k);
    for t=p+1:T
        x_err(t-p,:) = X(t,:) - a'*slice.dX(:,:,t-p);
        aa           = slice.dX(:,:,t-p)'*slice.mean.a_cov;
        XtilVXtil    = XtilVXtil + aa*slice.dX(:,:,t-p);
    end
    A = x_err'*x_err + XtilVXtil; % A in eq. 58/53 paper VB1 with completed square
end

P = lambda*A+diag(slice.mean.b);
C = inv(P);
d = diag(C);
R = C./sqrt(d*d');