%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% ND:
%   numerical differentiation - a equi-spaced one-dim array 'ind(.)' 
%   with step 'hstep'
% 
%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% By Dr Yangquan Chen (yqchen@ieee.org) 	
% Email=<yqchen@ieee.org>; URL=http://www.crosswinds.net/~yqchen/
% 07-07-1999. 										%
% Five-point numerical differentiating		%
%	using local cubic polynomial fit			%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
function [nd_outd]=nd5p(ind,hstep,npoints)
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% ND5P
nd_outd(1)=(-25.*ind(1)+48.*ind(2)-36.*ind(3)+16.*ind(4)- ...
   	3.*ind(5))/(12.*hstep);
nd_outd(2)=(-3.*ind(1)-10.*ind(2)+18.*ind(3)-6.*ind(4)+ ...
   	ind(5))/(12.*hstep);
for l=3:npoints-2
nd_outd(l)=(ind(l-2)-8.*ind(l-1)+8.*ind(l+1)-ind(l+2))/(12.*hstep);
end;
nd_outd(npoints-1)=(-ind(npoints-4)+6.*ind(npoints-3)- ...
		18.*ind(npoints-2)+10.*ind(npoints-1)+ ...
		3.*ind(npoints))/(12.*hstep);
nd_outd(npoints)=(3.*ind(npoints-4)-16.*ind(npoints-3) + ...
		36.*ind(npoints-2)-48.*ind(npoints-1)+ ...
   	25.*ind(npoints))/(12.*hstep);
return