Why not write this as an m file? I'm not sure what the symbolic toolbox buys you. 3 stars only because you used a toolbox I have no need for and no intention of buying.
26 Mar 2009
Find Optimal Kernel
Find optimal causal filter kernel that convolves with one time series to approximate another
Author: Matthew Nelson
for i = 1:r
Cxy(i,1)=InSig' * newOutSig(i+r-1,:);
Cxx(i,1)=InSig' * newInSig(i+r-1,:);
end
Cxy=Cxy(2:MaxnLags+1, :);
Cxx=Cxx(1:MaxnLags, :);
Cxxmat=toeplitz(Cxx);
k= (Cxxmat^-1)*Cxy;
25 Mar 2009
Find Optimal Kernel
Find optimal causal filter kernel that convolves with one time series to approximate another
Author: Matthew Nelson
% Without access to the Signal Processing Toolbox, it's difficult to % say for sure, but I suspect the following code would allow you to % run this routine without calling the xcov() function.
for i = 1:r
Cxy(i,1)=InSig' * newOutSig(i+r-1,:);
Cxx(i,1)=InSig' * newOutSig(i+r-1,:);
end
Cxy=Cxy(2:MaxnLags+1, :);
Cxx=Cxx(1:MaxnLags, :);
Cxxmat=toeplitz(Cxx);
k= (Cxxmat^-1)*Cxy;
25 Mar 2009
Find Optimal Kernel
Find optimal causal filter kernel that convolves with one time series to approximate another
Author: Matthew Nelson
25 Mar 2009
Find Optimal Kernel
Find optimal causal filter kernel that convolves with one time series to approximate another
Author: Matthew Nelson
Maybe you could provide a simple sample input & output so users can know they have the code working properly -- especially those of us who don't have the signal processing toolbox. Thx, N.
