Estimation optimal lag length VAR model (Matlab version from 2014)

11 views (last 30 days)
I have the following piece of code which estimates the optimal lag length for a VAR model, with the first part running on matlab versions which already include the estimate and varm functions and the latter part supposed to do the same in older versions. (I have remote access to a computer where the 2014a version is installed without being able to update it.) Currently, the results I am getting with the latter part (running on old Matlab) are very different from those using the newer functions. Is there a better way to produce the same results with the older Matlab version?
[~,d] = version;
for k=1:10
if year(d) >= 2017
Mdl = varm(2,k);
EstMdl = estimate(Mdl,Input);
Results = summarize(EstMdl);
BICtable(k) = Results.BIC;
else
Spec = vgxset('n',2,'nAR',k,'Constant',true);
[EstSpec, EstStdErrors,LLF] = vgxvarx(Spec, Input);
numParam = 2+ k*4;
nObs = size(Input,1);
[~,BICtable(k)] = aicbic(LLF, numParam, nObs);
end
end
[~, In2] = min(BICtable);
OptimLag = In2-1;

Answers (0)

Categories

Find more on Conditional Mean Models in Help Center and File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!