Dear all,
There seems to be a bug in dsge_likelihood.m that occurs when calling
estimation after identification. In the attached mod-file in
dsge_likelihood.m we have in lines 628 following:
if analytic_derivation==1,
analytic_deriv_info={analytic_derivation,DT,DYss,DOm,DH,DP,asy_Hess};
else
analytic_deriv_info={analytic_derivation,DT,DYss,DOm,DH,DP,D2T,D2Yss,D2Om,D2
H,D2P};
clear DT DYss DOm DH DP D2T D2Yss D2Om D2H D2P,
end
that is if analytic_derivation==2 set by identification we delete DH. But in
line 714 we access DH, which was deleted.
if analytic_derivation,
for j=1:pp,
tmp(j,:)=DH(j,j,:);
end
DH=tmp;
end
It seems we have to account for this case.
Best,
Johannes
--
Johannes Pfeifer
Haußerstr. 29
72076 Tübingen
Tel.: +49-(0)7071-6396184
Mobil.: +49-(0)170-6936820
Germany