function  A = m_gauss(b,X)
% A = m_gauss(b,X)
%
% The function returns the exponent values of X(:). The formula is:
% $$ A_j = exp(-\frac{(x-\mu_j)^2}{c_j}) $$
% The vector b contains n/2 groups of coefficients: mu(j) = b(2j-1) is a
% gaussian shift and c(j) = b(2j) relates to the width of the gaussian.
%
% Notation:
% X [m x 1] is an input column vector,
% b [1 x 2n] is a parameters vector, where n is a number of items in the
% non-linear mapping.
% A [m x n] is an output matrix of the non-linear  mappings, where each
% column corresponds to an item of the further linear combination. 
%
% Example
% y = m_gauss([20,0],linspace(0,1,10)) 
%
mu = b(1:2:end-1); % 2 parameters; even parameters are coeffs
c = b(2:2:end);   % 2 parameters; odd parameters are phase shifts
n = length(b)/2; % n components 
m = length(X);
C = repmat(1./c,[m,n]); % prepare a temporary matrix of denominators
A = exp( -((repmat(X(:),[1 n])).^2 ) .* C ); 
return