function y=fnorm(w,f,p)

%FNORM	Norms of MVFR matrix.

%	FNORM(W,F,p) applies NORM(Fm,p) to each component matrix

%	Fm of the MVFR matrix, F. The results are returned as

%	a column vector of length(W).

%	W is the associated frequency vector.

%	The input argument p is an optional parameter.

%	If p is missing the maximum singular values are returned.

%	For minimum singular values use

%	MIN(FSVD(W,F)')'

%

%	See NORM for further information and an explanation

%	of the possible values for the optional argument p.

%	NOTE: if the component matrices of F are vectors then

%	FNORM returns the vector results of NORM.



%	Dr M.P. Ford 4th August 1987

% Copyright (c) 1987 by GEC Engineering Research Centre & Cambridge Control Ltd



nargs=nargin;

error(nargchk(2,3,nargs));

[m,n]=fsize(w,f);

lw=length(w);

k=1:m;

y=zeros(lw,1);

if nargs==3

  for j=1:lw

    y(j)=norm(f(k+m*(j-1),:),p);  % break up f into matrices

  end

else

  for j=1:lw

    y(j,:)=norm(f(k+m*(j-1),:));  % break up f into matrices

  end

end