toon0554.m

老外写的小波变换的工具箱

% toon0554 -- AutoSpline Reconstructions from Noisy Data
%
%  Here we depict a reconstruction using penalized thin-plate splines,
%  in which periodic thin-plate splines are used to smooth the data,
%  with roughness penalty proportional to the energy stored in the
%  second derivative.  The constant of proportionality (or Lagrangian
%  tuning parameter) is the empirical minimizer of the Stein Unbiased
%  Estimate of Risk.
%
global yblocks ybumps yheavi yDoppler
global t
% 
clf;
SplineInit
%
	[xhat, c] = SplineUChoose(yblocks,2.);
	versaplot(221,t,xhat,[],' 4 (a) AutoSpline[Blocks]',[],[])
%
	[xhat,c] = SplineUChoose(ybumps,2.);
	versaplot(222,t,xhat,[],' 4 (b) AutoSpline[Bumps]',[],[])
%
	[xhat,c] = SplineUChoose(yheavi,2.);
	versaplot(223,t,xhat,[],' 4 (c) AutoSpline[HeaviSine]',[],[])
%
	[xhat,c] = SplineUChoose(yDoppler,2.);
	versaplot(224,t,xhat,[],' 4 (d) AutoSpline[Doppler]',[],[])
    
    
    
 
 
%
%  Part of Wavelab Version 850
%  Built Tue Jan  3 13:20:43 EST 2006
%  This is Copyrighted Material
%  For Copying permissions see COPYING.m
%  Comments? e-mail [e-mail]