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

Add flags to deal with errors on Cholesky matrices in CPF filter.

rm-particles^2
Frédéric Karamé 2018-06-11 09:46:17 +02:00
parent 1f08164b05
commit 43615ce4f6
2 changed files with 27 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
src/conditional_filter_proposal.m
View File

@ -1,4 +1,4 @@
function [ProposalStateVector,Weights] = conditional_filter_proposal(ReducedForm,obs,StateVectors,SampleWeights,Q_lower_triangular_cholesky,H_lower_triangular_cholesky,H,ParticleOptions,ThreadsOptions,normconst2)
function [ProposalStateVector,Weights,flag] = conditional_filter_proposal(ReducedForm,obs,StateVectors,SampleWeights,Q_lower_triangular_cholesky,H_lower_triangular_cholesky,H,ParticleOptions,ThreadsOptions,normconst2)
%
% Computes the proposal for each past particle using Gaussian approximations
% for the state errors and the Kalman filter
@ -42,6 +42,7 @@ persistent init_flag2 mf0 mf1
persistent number_of_state_variables number_of_observed_variables
persistent number_of_structural_innovations
flag=0 ;
% Set local state space model (first-order approximation).
ghx = ReducedForm.ghx;
ghu = ReducedForm.ghu;
@ -118,15 +119,31 @@ else
Error = obs - PredictedObservedMean ;
StateVectorMean = PredictedStateMean + KalmanFilterGain*Error ;
StateVectorVariance = PredictedStateVariance - KalmanFilterGain*PredictedObservedVariance*KalmanFilterGain';
StateVectorVarianceSquareRoot = chol(StateVectorVariance + eye(number_of_state_variables)*1e-6)' ;
StateVectorVariance = 0.5*(StateVectorVariance+StateVectorVariance');
%StateVectorVarianceSquareRoot = reduced_rank_cholesky(StateVectorVariance)';%chol(StateVectorVariance + eye(number_of_state_variables)*1e-6)' ;
[StateVectorVarianceSquareRoot,p] = chol(StateVectorVariance,'lower') ;
if p
flag=1;
ProposalStateVector = zeros(number_of_state_variables,1) ;
Weights = 0.0 ;
return
end
if ParticleOptions.cpf_weights_method.amisanotristani
Weights = SampleWeights.*probability2(zeros(number_of_observed_variables,1),chol(PredictedObservedVariance)',Error) ;
end
end
PredictedStateVarianceSquareRoot = chol(PredictedStateVariance + eye(number_of_state_variables)*1e-6)' ;
ProposalStateVector = StateVectorVarianceSquareRoot*randn(size(StateVectorVarianceSquareRoot,2),1)+StateVectorMean ;
if ParticleOptions.cpf_weights_method.murrayjonesparslow
PredictedStateVariance = 0.5*(PredictedStateVariance+PredictedStateVariance');
%PredictedStateVarianceSquareRoot = reduced_rank_cholesky(PredictedStateVariance)';%chol(PredictedStateVariance + eye(number_of_state_variables)*1e-6)' ;
[PredictedStateVarianceSquareRoot,p] = chol(PredictedStateVariance,'lower') ;
if p
flag=1;
ProposalStateVector = zeros(number_of_state_variables,1) ;
Weights = 0.0 ;
return
end
Prior = probability2(PredictedStateMean,PredictedStateVarianceSquareRoot,ProposalStateVector) ;
Posterior = probability2(StateVectorMean,StateVectorVarianceSquareRoot,ProposalStateVector) ;
Likelihood = probability2(obs,H_lower_triangular_cholesky,measurement_equations(ProposalStateVector,ReducedForm,ThreadsOptions)) ;

9
src/conditional_particle_filter.m
View File

@ -103,8 +103,13 @@ StateParticles = bsxfun(@plus,StateVectorVarianceSquareRoot*randn(state_variance
SampleWeights = ones(1,number_of_particles)/number_of_particles ;
for t=1:sample_size
for i=1:number_of_particles
[StateParticles(:,i),SampleWeights(i)] = ...
[StateParticles(:,i),SampleWeights(i),flag] = ...
conditional_filter_proposal(ReducedForm,Y(:,t),StateParticles(:,i),SampleWeights(i),Q_lower_triangular_cholesky,H_lower_triangular_cholesky,H,ParticleOptions,ThreadsOptions,normconst2) ;
if flag==1
LIK=-Inf;
lik(t)=-Inf;
return
end
end
SumSampleWeights = sum(SampleWeights) ;
lik(t) = log(SumSampleWeights) ;
@ -116,4 +121,4 @@ for t=1:sample_size
end
end
LIK = -sum(lik(start:end));
LIK = -sum(lik(start:end));