Replaced particle folder by a Git submodule (particles). Adapted path in dynare_config.

parent 7b90fd9c
......@@ -13,3 +13,6 @@
[submodule "matlab/utilities/tests"]
path = matlab/utilities/tests
url = https://github.com/DynareTeam/m-unit-tests.git
[submodule "matlab/particles"]
path = matlab/particles
url = https://github.com/DynareTeam/particles.git
......@@ -44,7 +44,6 @@ if ~nargin || nargin==1
verbose = 1;
end
addpath([dynareroot '/distributions/'])
addpath([dynareroot '/kalman/'])
addpath([dynareroot '/kalman/likelihood'])
......@@ -54,7 +53,7 @@ addpath([dynareroot '/ms-sbvar/'])
addpath([dynareroot '/ms-sbvar/identification/'])
addpath([dynareroot '../contrib/ms-sbvar/TZcode/MatlabFiles/'])
addpath([dynareroot '/parallel/'])
addpath([dynareroot '/particle/'])
addpath([dynareroot '/particles/src'])
addpath([dynareroot '/gsa/'])
addpath([dynareroot '/ep/'])
addpath([dynareroot '/lmmcp/'])
......
function initial_distribution = auxiliary_initialization(ReducedForm,Y,start,DynareOptions)
% Evaluates the likelihood of a nonlinear model with a particle filter allowing eventually resampling.
%
% INPUTS
% ReducedForm [structure] Matlab's structure describing the reduced form model.
% ReducedForm.measurement.H [double] (pp x pp) variance matrix of measurement errors.
% ReducedForm.state.Q [double] (qq x qq) variance matrix of state errors.
% ReducedForm.state.dr [structure] output of resol.m.
% Y [double] pp*smpl matrix of (detrended) data, where pp is the maximum number of observed variables.
% start [integer] scalar, likelihood evaluation starts at 'start'.
% mf [integer] pp*1 vector of indices.
% number_of_particles [integer] scalar.
%
% OUTPUTS
% LIK [double] scalar, likelihood
% lik [double] vector, density of observations in each period.
%
% REFERENCES
%
% NOTES
% The vector "lik" is used to evaluate the jacobian of the likelihood.
% Copyright (C) 2013 Dynare Team
%
% This file is part of Dynare.
%
% Dynare is free software: you can redistribute it and/or modify
% it under the terms of the GNU General Public License as published by
% the Free Software Foundation, either version 3 of the License, or
% (at your option) any later version.
%
% Dynare is distributed in the hope that it will be useful,
% but WITHOUT ANY WARRANTY; without even the implied warranty of
% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
% GNU General Public License for more details.
%
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
persistent init_flag mf0 mf1 number_of_particles
persistent number_of_observed_variables number_of_structural_innovations
% Set default
if isempty(start)
start = 1;
end
% Set flag for prunning
%pruning = DynareOptions.particle.pruning;
% Get steady state and mean.
%steadystate = ReducedForm.steadystate;
constant = ReducedForm.constant;
state_variables_steady_state = ReducedForm.state_variables_steady_state;
% Set persistent variables.
if isempty(init_flag)
mf0 = ReducedForm.mf0;
mf1 = ReducedForm.mf1;
number_of_observed_variables = length(mf1);
number_of_structural_innovations = length(ReducedForm.Q);
number_of_particles = DynareOptions.particle.number_of_particles;
init_flag = 1;
end
% Set local state space model (first order approximation).
ghx = ReducedForm.ghx;
ghu = ReducedForm.ghu;
% Set local state space model (second order approximation).
ghxx = ReducedForm.ghxx;
ghuu = ReducedForm.ghuu;
ghxu = ReducedForm.ghxu;
% Get covariance matrices
Q = ReducedForm.Q;
H = ReducedForm.H;
if isempty(H)
H = 0;
end
% Get initial condition for the state vector.
StateVectorMean = ReducedForm.StateVectorMean;
StateVectorVarianceSquareRoot = reduced_rank_cholesky(ReducedForm.StateVectorVariance)';
state_variance_rank = size(StateVectorVarianceSquareRoot,2);
%Q_lower_triangular_cholesky = chol(Q)';
%if pruning
% StateVectorMean_ = StateVectorMean;
% StateVectorVarianceSquareRoot_ = StateVectorVarianceSquareRoot;
%end
% Set seed for randn().
set_dynare_seed('default');
% Initialization of the likelihood.
const_lik = log(2*pi)*number_of_observed_variables;
% Initialization of the weights across particles.
weights = ones(1,number_of_particles)/number_of_particles ;
StateVectors = bsxfun(@plus,StateVectorVarianceSquareRoot*randn(state_variance_rank,number_of_particles),StateVectorMean);
%if pruning
% StateVectors_ = StateVectors;
%end
yhat = bsxfun(@minus,StateVectors,state_variables_steady_state);
%if pruning
% yhat_ = bsxfun(@minus,StateVectors_,state_variables_steady_state);
% [tmp, tmp_] = local_state_space_iteration_2(yhat,zeros(number_of_structural_innovations,number_of_particles),ghx,ghu,constant,ghxx,ghuu,ghxu,yhat_,steadystate,DynareOptions.threads.local_state_space_iteration_2);
%else
tmp = local_state_space_iteration_2(yhat,zeros(number_of_structural_innovations,number_of_particles),ghx,ghu,constant,ghxx,ghuu,ghxu,DynareOptions.threads.local_state_space_iteration_2);
%end
PredictedObservedMean = weights*(tmp(mf1,:)');
PredictionError = bsxfun(@minus,Y(:,t),tmp(mf1,:));
dPredictedObservedMean = bsxfun(@minus,tmp(mf1,:),PredictedObservedMean');
PredictedObservedVariance = bsxfun(@times,weights,dPredictedObservedMean)*dPredictedObservedMean' + H;
wtilde = exp(-.5*(const_lik+log(det(PredictedObservedVariance))+sum(PredictionError.*(PredictedObservedVariance\PredictionError),1))) ;
tau_tilde = weights.*wtilde ;
tau_tilde = tau_tilde/sum(tau_tilde);
initial_distribution = resample(StateVectors',tau_tilde',DynareOptions)' ;
\ No newline at end of file
function [LIK,lik] = auxiliary_particle_filter(ReducedForm,Y,start,DynareOptions)
% Evaluates the likelihood of a nonlinear model with a particle filter allowing eventually resampling.
%
% INPUTS
% ReducedForm [structure] Matlab's structure describing the reduced form model.
% ReducedForm.measurement.H [double] (pp x pp) variance matrix of measurement errors.
% ReducedForm.state.Q [double] (qq x qq) variance matrix of state errors.
% ReducedForm.state.dr [structure] output of resol.m.
% Y [double] pp*smpl matrix of (detrended) data, where pp is the maximum number of observed variables.
% start [integer] scalar, likelihood evaluation starts at 'start'.
% mf [integer] pp*1 vector of indices.
% number_of_particles [integer] scalar.
%
% OUTPUTS
% LIK [double] scalar, likelihood
% lik [double] vector, density of observations in each period.
%
% REFERENCES
%
% NOTES
% The vector "lik" is used to evaluate the jacobian of the likelihood.
% Copyright (C) 2011-2013 Dynare Team
%
% This file is part of Dynare.
%
% Dynare is free software: you can redistribute it and/or modify
% it under the terms of the GNU General Public License as published by
% the Free Software Foundation, either version 3 of the License, or
% (at your option) any later version.
%
% Dynare is distributed in the hope that it will be useful,
% but WITHOUT ANY WARRANTY; without even the implied warranty of
% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
% GNU General Public License for more details.
%
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
persistent init_flag mf0 mf1 number_of_particles
persistent sample_size number_of_state_variables number_of_observed_variables number_of_structural_innovations
% Set default
if isempty(start)
start = 1;
end
% Set flag for prunning
pruning = DynareOptions.particle.pruning;
% Get steady state and mean.
steadystate = ReducedForm.steadystate;
constant = ReducedForm.constant;
state_variables_steady_state = ReducedForm.state_variables_steady_state;
% Set persistent variables.
if isempty(init_flag)
mf0 = ReducedForm.mf0;
mf1 = ReducedForm.mf1;
sample_size = size(Y,2);
number_of_state_variables = length(mf0);
number_of_observed_variables = length(mf1);
number_of_structural_innovations = length(ReducedForm.Q);
number_of_particles = DynareOptions.particle.number_of_particles;
init_flag = 1;
end
% Set local state space model (first order approximation).
ghx = ReducedForm.ghx;
ghu = ReducedForm.ghu;
% Set local state space model (second order approximation).
ghxx = ReducedForm.ghxx;
ghuu = ReducedForm.ghuu;
ghxu = ReducedForm.ghxu;
% Get covariance matrices
Q = ReducedForm.Q;
H = ReducedForm.H;
if isempty(H)
H = 0;
end
% Get initial condition for the state vector.
StateVectorMean = ReducedForm.StateVectorMean;
StateVectorVarianceSquareRoot = reduced_rank_cholesky(ReducedForm.StateVectorVariance)';
state_variance_rank = size(StateVectorVarianceSquareRoot,2);
Q_lower_triangular_cholesky = chol(Q)';
if pruning
StateVectorMean_ = StateVectorMean;
StateVectorVarianceSquareRoot_ = StateVectorVarianceSquareRoot;
end
% Set seed for randn().
set_dynare_seed('default');
% Initialization of the likelihood.
const_lik = log(2*pi)*number_of_observed_variables;
lik = NaN(sample_size,1);
LIK = NaN;
% Initialization of the weights across particles.
weights = ones(1,number_of_particles)/number_of_particles ;
StateVectors = bsxfun(@plus,StateVectorVarianceSquareRoot*randn(state_variance_rank,number_of_particles),StateVectorMean);
if pruning
StateVectors_ = StateVectors;
end
for t=1:sample_size
yhat = bsxfun(@minus,StateVectors,state_variables_steady_state);
if pruning
yhat_ = bsxfun(@minus,StateVectors_,state_variables_steady_state);
[tmp, tmp_] = local_state_space_iteration_2(yhat,zeros(number_of_structural_innovations,number_of_particles),ghx,ghu,constant,ghxx,ghuu,ghxu,yhat_,steadystate,DynareOptions.threads.local_state_space_iteration_2);
else
tmp = local_state_space_iteration_2(yhat,zeros(number_of_structural_innovations,number_of_particles),ghx,ghu,constant,ghxx,ghuu,ghxu,DynareOptions.threads.local_state_space_iteration_2);
end
PredictedObservedMean = weights*(tmp(mf1,:)');
PredictionError = bsxfun(@minus,Y(:,t),tmp(mf1,:));
dPredictedObservedMean = bsxfun(@minus,tmp(mf1,:),PredictedObservedMean');
PredictedObservedVariance = bsxfun(@times,weights,dPredictedObservedMean)*dPredictedObservedMean' +H;
wtilde = exp(-.5*(const_lik+log(det(PredictedObservedVariance))+sum(PredictionError.*(PredictedObservedVariance\PredictionError),1))) ;
tau_tilde = weights.*wtilde ;
sum_tau_tilde = sum(tau_tilde) ;
%var_wtilde = wtilde-sum_tau_tilde ;
%var_wtilde = var_wtilde'*var_wtilde/(number_of_particles-1) ;
lik(t) = log(sum_tau_tilde) ; %+ .5*var_wtilde/(number_of_particles*(sum_tau_tilde*sum_tau_tilde)) ;
tau_tilde = tau_tilde/sum_tau_tilde;
if pruning
temp = resample([yhat' yhat_'],tau_tilde',DynareOptions);
yhat = temp(:,1:number_of_state_variables)' ;
yhat_ = temp(:,number_of_state_variables+1:2*number_of_state_variables)' ;
else
yhat = resample(yhat',tau_tilde',DynareOptions)' ;
end
if pruning
[tmp, tmp_] = local_state_space_iteration_2(yhat,zeros(number_of_structural_innovations,number_of_particles),ghx,ghu,constant,ghxx,ghuu,ghxu,yhat_,steadystate,DynareOptions.threads.local_state_space_iteration_2);
else
tmp = local_state_space_iteration_2(yhat,zeros(number_of_structural_innovations,number_of_particles),ghx,ghu,constant,ghxx,ghuu,ghxu,DynareOptions.threads.local_state_space_iteration_2);
end
PredictedObservedMean = weights*(tmp(mf1,:)');
PredictionError = bsxfun(@minus,Y(:,t),tmp(mf1,:));
dPredictedObservedMean = bsxfun(@minus,tmp(mf1,:),PredictedObservedMean');
PredictedObservedVariance = bsxfun(@times,weights,dPredictedObservedMean)*dPredictedObservedMean' +H;
wtilde = exp(-.5*(const_lik+log(det(PredictedObservedVariance))+sum(PredictionError.*(PredictedObservedVariance\PredictionError),1))) ;
epsilon = Q_lower_triangular_cholesky*randn(number_of_structural_innovations,number_of_particles);
if pruning
[tmp, tmp_] = local_state_space_iteration_2(yhat,epsilon,ghx,ghu,constant,ghxx,ghuu,ghxu,yhat_,steadystate,DynareOptions.threads.local_state_space_iteration_2);
StateVectors_ = tmp_(mf0,:);
else
tmp = local_state_space_iteration_2(yhat,epsilon,ghx,ghu,constant,ghxx,ghuu,ghxu,DynareOptions.threads.local_state_space_iteration_2);
end
StateVectors = tmp(mf0,:);
PredictedObservedMean = mean(tmp(mf1,:),2);
PredictionError = bsxfun(@minus,Y(:,t),tmp(mf1,:));
dPredictedObservedMean = bsxfun(@minus,tmp(mf1,:),PredictedObservedMean);
PredictedObservedVariance = (dPredictedObservedMean*dPredictedObservedMean')/number_of_particles + H;
lnw = exp(-.5*(const_lik+log(det(PredictedObservedVariance))+sum(PredictionError.*(PredictedObservedVariance\PredictionError),1)));
wtilde = lnw./wtilde;
weights = wtilde/sum(wtilde);
end
LIK = -sum(lik(start:end));
\ No newline at end of file
function [ProposalStateVector,Weights] = conditional_filter_proposal(ReducedForm,obs,StateVectors,SampleWeights,Q_lower_triangular_cholesky,H_lower_triangular_cholesky,H,DynareOptions,normconst2)
%
% Computes the proposal for each past particle using Gaussian approximations
% for the state errors and the Kalman filter
%
% INPUTS
% reduced_form_model [structure] Matlab's structure describing the reduced form model.
% reduced_form_model.measurement.H [double] (pp x pp) variance matrix of measurement errors.
% reduced_form_model.state.Q [double] (qq x qq) variance matrix of state errors.
% reduced_form_model.state.dr [structure] output of resol.m.
% Y [double] pp*smpl matrix of (detrended) data, where pp is the maximum number of observed variables.
%
% OUTPUTS
% LIK [double] scalar, likelihood
% lik [double] vector, density of observations in each period.
%
% REFERENCES
%
% NOTES
% The vector "lik" is used to evaluate the jacobian of the likelihood.
% Copyright (C) 2012-2013 Dynare Team
%
% This file is part of Dynare.
%
% Dynare is free software: you can redistribute it and/or modify
% it under the terms of the GNU General Public License as published by
% the Free Software Foundation, either version 3 of the License, or
% (at your option) any later version.
%
% Dynare is distributed in the hope that it will be useful,
% but WITHOUT ANY WARRANTY; without even the implied warranty of
% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
% GNU General Public License for more details.
%
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
%
% AUTHOR(S) frederic DOT karame AT univ DASH lemans DOT fr
% stephane DOT adjemian AT univ DASH lemans DOT fr
persistent init_flag2 mf0 mf1
persistent number_of_state_variables number_of_observed_variables
persistent number_of_structural_innovations
% Set local state space model (first-order approximation).
ghx = ReducedForm.ghx;
ghu = ReducedForm.ghu;
% Set local state space model (second-order approximation).
ghxx = ReducedForm.ghxx;
ghuu = ReducedForm.ghuu;
ghxu = ReducedForm.ghxu;
if any(any(isnan(ghx))) || any(any(isnan(ghu))) || any(any(isnan(ghxx))) || any(any(isnan(ghuu))) || any(any(isnan(ghxu))) || ...
any(any(isinf(ghx))) || any(any(isinf(ghu))) || any(any(isinf(ghxx))) || any(any(isinf(ghuu))) || any(any(isinf(ghxu))) ...
any(any(abs(ghx)>1e4)) || any(any(abs(ghu)>1e4)) || any(any(abs(ghxx)>1e4)) || any(any(abs(ghuu)>1e4)) || any(any(abs(ghxu)>1e4))
ghx
ghu
ghxx
ghuu
ghxu
end
constant = ReducedForm.constant;
state_variables_steady_state = ReducedForm.state_variables_steady_state;
% Set persistent variables.
if isempty(init_flag2)
mf0 = ReducedForm.mf0;
mf1 = ReducedForm.mf1;
number_of_state_variables = length(mf0);
number_of_observed_variables = length(mf1);
number_of_structural_innovations = length(ReducedForm.Q);
init_flag2 = 1;
end
if DynareOptions.particle.proposal_approximation.cubature || DynareOptions.particle.proposal_approximation.montecarlo
[nodes,weights] = spherical_radial_sigma_points(number_of_structural_innovations);
weights_c = weights ;
elseif DynareOptions.particle.proposal_approximation.unscented
[nodes,weights,weights_c] = unscented_sigma_points(number_of_structural_innovations,DynareOptions);
else
error('Estimation: This approximation for the proposal is not implemented or unknown!')
end
epsilon = Q_lower_triangular_cholesky*(nodes') ;
yhat = repmat(StateVectors-state_variables_steady_state,1,size(epsilon,2)) ;
tmp = local_state_space_iteration_2(yhat,epsilon,ghx,ghu,constant,ghxx,ghuu,ghxu,DynareOptions.threads.local_state_space_iteration_2);
PredictedStateMean = tmp(mf0,:)*weights ;
PredictedObservedMean = tmp(mf1,:)*weights;
if DynareOptions.particle.proposal_approximation.cubature || DynareOptions.particle.proposal_approximation.montecarlo
PredictedStateMean = sum(PredictedStateMean,2) ;
PredictedObservedMean = sum(PredictedObservedMean,2) ;
dState = bsxfun(@minus,tmp(mf0,:),PredictedStateMean)'.*sqrt(weights) ;
dObserved = bsxfun(@minus,tmp(mf1,:),PredictedObservedMean)'.*sqrt(weights);
big_mat = [dObserved dState; [H_lower_triangular_cholesky zeros(number_of_observed_variables,number_of_state_variables)] ];
[mat1,mat] = qr2(big_mat,0);
mat = mat';
clear('mat1');
PredictedObservedVarianceSquareRoot = mat(1:number_of_observed_variables,1:number_of_observed_variables);
CovarianceObservedStateSquareRoot = mat(number_of_observed_variables+(1:number_of_state_variables),1:number_of_observed_variables);
StateVectorVarianceSquareRoot = mat(number_of_observed_variables+(1:number_of_state_variables),number_of_observed_variables+(1:number_of_state_variables));
StateVectorMean = PredictedStateMean + (CovarianceObservedStateSquareRoot/PredictedObservedVarianceSquareRoot)*(obs - PredictedObservedMean);
else
dState = bsxfun(@minus,tmp(mf0,:),PredictedStateMean);
dObserved = bsxfun(@minus,tmp(mf1,:),PredictedObservedMean);
PredictedStateVariance = dState*diag(weights_c)*dState';
PredictedObservedVariance = dObserved*diag(weights_c)*dObserved' + H;
PredictedStateAndObservedCovariance = dState*diag(weights_c)*dObserved';
KalmanFilterGain = PredictedStateAndObservedCovariance/PredictedObservedVariance ;
StateVectorMean = PredictedStateMean + KalmanFilterGain*(obs - PredictedObservedMean);
StateVectorVariance = PredictedStateVariance - KalmanFilterGain*PredictedObservedVariance*KalmanFilterGain';
StateVectorVariance = .5*(StateVectorVariance+StateVectorVariance');
StateVectorVarianceSquareRoot = reduced_rank_cholesky(StateVectorVariance)';
end
ProposalStateVector = StateVectorVarianceSquareRoot*randn(size(StateVectorVarianceSquareRoot,2),1)+StateVectorMean ;
ypred = measurement_equations(ProposalStateVector,ReducedForm,DynareOptions) ;
foo = H_lower_triangular_cholesky \ (obs - ypred) ;
likelihood = exp(-0.5*(foo)'*foo)/normconst2 + 1e-99 ;
Weights = SampleWeights.*likelihood;
function [LIK,lik] = conditional_particle_filter(ReducedForm,Y,start,DynareOptions)
%
% Evaluates the likelihood of a non-linear model with a particle filter
% - the proposal is built using the Kalman updating step for each particle.
% - we need draws in the errors distributions
% Whether we use Monte-Carlo draws from a multivariate gaussian distribution
% as in Amisano & Tristani (JEDC 2010).
% Whether we use multidimensional Gaussian sparse grids approximations:
% - a univariate Kronrod-Paterson Gaussian quadrature combined by the Smolyak
% operator (ref: Winschel & Kratzig, 2010).
% - a spherical-radial cubature (ref: Arasaratnam & Haykin, 2009a,2009b).
% - a scaled unscented transform cubature (ref: Julier & Uhlmann 1997, van der
% Merwe & Wan 2003).
%
% Pros:
% - Allows using current observable information in the proposal
% - The use of sparse grids Gaussian approximation is much faster than the Monte-Carlo approach
% Cons:
% - The use of the Kalman updating step may biais the proposal distribution since
% it has been derived in a linear context and is implemented in a nonlinear
% context. That is why particle resampling is performed.
%
% INPUTS
% reduced_form_model [structure] Matlab's structure describing the reduced form model.
% reduced_form_model.measurement.H [double] (pp x pp) variance matrix of measurement errors.
% reduced_form_model.state.Q [double] (qq x qq) variance matrix of state errors.
% reduced_form_model.state.dr [structure] output of resol.m.
% Y [double] pp*smpl matrix of (detrended) data, where pp is the maximum number of observed variables.
% start [integer] scalar, likelihood evaluation starts at 'start'.
% smolyak_accuracy [integer] scalar.
%
% OUTPUTS
% LIK [double] scalar, likelihood
% lik [double] vector, density of observations in each period.
%
% REFERENCES
%
% NOTES
% The vector "lik" is used to evaluate the jacobian of the likelihood.
% Copyright (C) 2009-2010 Dynare Team
%
% This file is part of Dynare.
%
% Dynare is free software: you can redistribute it and/or modify
% it under the terms of the GNU General Public License as published by
% the Free Software Foundation, either version 3 of the License, or
% (at your option) any later version.
%
% Dynare is distributed in the hope that it will be useful,
% but WITHOUT ANY WARRANTY; without even the implied warranty of
% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
% GNU General Public License for more details.
%
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
% AUTHOR(S) frederic DOT karame AT univ DASH lemans DOT fr
% stephane DOT adjemian AT univ DASH lemans DOT fr
persistent init_flag mf0 mf1
persistent number_of_particles
persistent sample_size number_of_state_variables number_of_observed_variables
% Set default
if isempty(start)
start = 1;
end
% Set persistent variables.
if isempty(init_flag)
mf0 = ReducedForm.mf0;
mf1 = ReducedForm.mf1;
sample_size = size(Y,2);
number_of_state_variables = length(mf0);
number_of_observed_variables = length(mf1);
init_flag = 1;
number_of_particles = DynareOptions.particle.number_of_particles ;
end
% Get covariance matrices
Q = ReducedForm.Q;
H = ReducedForm.H;
if isempty(H)
H = 0;
H_lower_triangular_cholesky = 0;
else
H_lower_triangular_cholesky = reduced_rank_cholesky(H)';
end
% Get initial condition for the state vector.
StateVectorMean = ReducedForm.StateVectorMean;
StateVectorVarianceSquareRoot = reduced_rank_cholesky(ReducedForm.StateVectorVariance)';
state_variance_rank = size(StateVectorVarianceSquareRoot,2);
Q_lower_triangular_cholesky = reduced_rank_cholesky(Q)';
% Set seed for randn().
set_dynare_seed('default');
% Initialization of the likelihood.
normconst2 = log(2*pi)*number_of_observed_variables*prod(diag(H_lower_triangular_cholesky)) ;
lik = NaN(sample_size,1);
LIK = NaN;
ks = 0 ;
StateParticles = bsxfun(@plus,StateVectorVarianceSquareRoot*randn(state_variance_rank,number_of_particles),StateVectorMean);
SampleWeights = ones(1,number_of_particles)/number_of_particles ;
for t=1:sample_size
for i=1:number_of_particles
[StateParticles(:,i),SampleWeights(i)] = ...
conditional_filter_proposal(ReducedForm,Y(:,t),StateParticles(:,i),SampleWeights(i),Q_lower_triangular_cholesky,H_lower_triangular_cholesky,H,DynareOptions,normconst2) ;
end
SumSampleWeights = sum(SampleWeights) ;
lik(t) = log(SumSampleWeights) ;
SampleWeights = SampleWeights./SumSampleWeights ;
if (DynareOptions.particle.resampling.status.generic && neff(SampleWeights)<DynareOptions.particle.resampling.threshold*sample_size) || DynareOptions.particle.resampling.status.systematic
ks = ks + 1 ;
StateParticles = resample(StateParticles',SampleWeights',DynareOptions)';
SampleWeights = ones(1,number_of_particles)/number_of_particles ;
end
end
LIK = -sum(lik(start:end));
function [StateMu,StateSqrtP,StateWeights] = fit_gaussian_mixture(X,StateMu,StateSqrtP,StateWeights,crit,niters,check)
% Copyright (C) 2013 Dynare Team
%
% This file is part of Dynare.
%
% Dynare is free software: you can redistribute it and/or modify
% it under the terms of the GNU General Public License as published by
% the Free Software Foundation, either version 3 of the License, or
% (at your option) any later version.
%
% Dynare is distributed in the hope that it will be useful,
% but WITHOUT ANY WARRANTY; without even the implied warranty of
% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
% GNU General Public License for more details.
%
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
[dim,Ndata] = size(X);
M = size(StateMu,2) ;
if check % Ensure that covariances don't collapse
MIN_COVAR_SQRT = sqrt(eps);
init_covars = StateSqrtP;
end
eold = -Inf;
for n=1:niters
% Calculate posteriors based on old parameters
[prior,likelihood,marginal,posterior] = probability(StateMu,StateSqrtP,StateWeights,X);
e = sum(log(marginal));
if (n > 1 && abs((e - eold)/eold) < crit)
return;
else
eold = e;
end
new_pr = (sum(posterior,2))';
StateWeights = new_pr/Ndata;
StateMu = bsxfun(@rdivide,(posterior*X')',new_pr);
for j=1:M
diffs = bsxfun(@minus,X,StateMu(:,j));
tpost = (1/sqrt(new_pr(j)))*sqrt(posterior(j,:));
diffs = bsxfun(@times,diffs,tpost);
[foo,tcov] = qr2(diffs',0);
StateSqrtP(:,:,j) = tcov';
if check
if min(abs(diag(StateSqrtP(:,:,j)))) < MIN_COVAR_SQRT
StateSqrtP(:,:,j) = init_covars(:,:,j);
end
end
end
end