function retval = leastsqhess(x)
% usage:  y = leastsqhess(x)
% description: returns least squares
% function gradient

% no bulletproofing
retval = leastsqjac(x);
% the casual approach .. this is just the least squares approxn.
retval = retval'*retval;
% the cautions approach, useful if ill conditioning is an issue
%[U,S,V] = svd(retval,0);
%retval = V*diag(diag(S).^2)*V';