stepan
/
dynare
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

v4.1:/tests/kalman Bug corrections. 

git-svn-id: https://www.dynare.org/svn/dynare/dynare_v4@2175 ac1d8469-bf42-47a9-8791-bf33cf982152
time-shift
adjemian 2008-10-17 16:22:25 +00:00
parent 8aab45ce5d
commit afdddb8091
2 changed files with 14 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
tests/kalman/likelihood/compare_kalman_routines.m
View File

@ -43,6 +43,8 @@ end
% I randomly choose the mm eigenvalues of the transition matrix.
TransitionEigenvalues = rand(mm,1)*2-1;
% I randomly choose the mm eigenvectors of the transition matrix
tmp = rand(mm,mm*100);
TransitionEigenvectors = tmp*tmp'/(mm*100);
TransitionEigenvectors = rand(mm,mm);
% I build the transition matrix
T = TransitionEigenvectors*diag(TransitionEigenvalues)*inv(TransitionEigenvectors);
@ -68,6 +70,7 @@ mf = MF(1:pp);
P = lyapunov_symm(T,R*Q*R',1.000001);
if PeriodsWithMissingObservations==0
% kalman_filter.m
@ -79,7 +82,7 @@ if PeriodsWithMissingObservations==0
HH = H;
end
instant0 = clock;
LIK1 = kalman_filter(T,R,Q,HH,P,Y,start,mf,kalman_tol,riccati_tol);
[LIK1,lik1] = kalman_filter(T,R,Q,HH,P,Y,start,mf,kalman_tol,riccati_tol);
T1 = etime(clock, instant0);
disp(['kalman_filter = ' num2str(T1)])
@ -92,7 +95,7 @@ if PeriodsWithMissingObservations==0
HH = H;
end
instant0 = clock;
LIK2 = missing_observations_kalman_filter(T,R,Q,HH,P,Y,start,mf,kalman_tol,riccati_tol,data_index,number_of_observations,no_more_missing_observations);
[LIK2,lik2] = missing_observations_kalman_filter(T,R,Q,HH,P,Y,start,mf,kalman_tol,riccati_tol,data_index,number_of_observations,no_more_missing_observations);
T2 = etime(clock, instant0);
disp(['missing_observations_kalman_filter = ' num2str(T2)])
@ -106,24 +109,25 @@ if PeriodsWithMissingObservations==0
error('The univariate approach cannot handle correlated measurement errors')
end
instant0 = clock;
LIK3 = univariate_kalman_filter(T,R,Q,HH,P,Y,start,mf,kalman_tol,riccati_tol,data_index,number_of_observations,no_more_missing_observations);
[LIK3,lik3] = univariate_kalman_filter(T,R,Q,HH,P,Y,start,mf,kalman_tol,riccati_tol,data_index,number_of_observations,no_more_missing_observations);
T3 = etime(clock, instant0);
disp(['univariate_kalman_filter = ' num2str(T2)])
disp(['univariate_kalman_filter = ' num2str(T3)])
disp(' ')
disp(' ')
if abs(LIK1-LIK2)<1e-12
if abs(LIK2/LIK1-1)<1e-9
disp('missing data version is Ok')
else
disp('missing data version is wrong')
LIK1-LIK2
disp(['percentage dev. = ' num2str((LIK2/LIK1-1)*100)])
end
if abs(LIK1-LIK3)<1e-15
if abs(LIK3/LIK1-1)<1e-9
disp('univariate version is Ok')
else
disp('univariate version is wrong')
LIK1-LIK3
disp(['percentage dev. = ' num2str((LIK3/LIK1-1)*100)])
[lik1/lik1(end),lik3/lik3(end)]
end
else
% missing_observations_kalman_filter.m

4
tests/kalman/likelihood/test1.m
View File

@ -1,8 +1,8 @@
Experience.Number0fObservedVariables = 10;
Experience.SizeOfTheStateVector = 50;
Experience.SizeOfTheStateVector = 100;
Experience.NumberOfStructuralShocks = 12;
Experience.MeasurementErrors = 0;
Experience.NumberOfPeriods = 100;
Experience.NumberOfPeriods = 190;
Experience.PercentageOfMissingObservations = 0;
Experience.PeriodsWithMissingObservations = 0;