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

Added new functions for posterior distribution processing (second order moments) + Removed global variables. 

git-svn-id: https://www.dynare.org/svn/dynare/dynare_v4@1890 ac1d8469-bf42-47a9-8791-bf33cf982152
time-shift
adjemian 2008-06-21 08:33:31 +00:00
parent e315e73577
commit 341adad2e9
26 changed files with 416 additions and 265 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
matlab/DsgeLikelihood.m
View File

@ -1,5 +1,4 @@
function [fval,cost_flag,ys,trend_coeff,info] = DsgeLikelihood(xparam1,gend,data)
% function [fval,cost_flag,ys,trend_coeff,info] = DsgeLikelihood(xparam1,gend,data)
% Evaluates the posterior kernel of a dsge model.
%
@ -22,9 +21,7 @@ function [fval,cost_flag,ys,trend_coeff,info] = DsgeLikelihood(xparam1,gend,data
%
% part of DYNARE, copyright Dynare Team (2004-2008)
% Gnu Public License.
global bayestopt_ estim_params_ options_ trend_coeff_ M_ oo_ xparam1_test
fval = [];
ys = [];
trend_coeff = [];
@ -155,7 +152,7 @@ function [fval,cost_flag,ys,trend_coeff,info] = DsgeLikelihood(xparam1,gend,data
% 3. Initial condition of the Kalman filter
%------------------------------------------------------------------------------
if options_.lik_init == 1 % Kalman filter
Pstar = lyapunov_symm(T,R*Q*R');
Pstar = lyapunov_symm(T,R*Q*R',options_.qz_criterium);
Pinf = [];
elseif options_.lik_init == 2 % Old Diffuse Kalman filter
Pstar = 10*eye(np);
@ -165,7 +162,7 @@ function [fval,cost_flag,ys,trend_coeff,info] = DsgeLikelihood(xparam1,gend,data
Pstar = zeros(np,np);
ivs = bayestopt_.restrict_var_list_stationary;
R1 = R(ivs,:);
Pstar(ivs,ivs) = lyapunov_symm(T(ivs,ivs),R1*Q*R1');
Pstar(ivs,ivs) = lyapunov_symm(T(ivs,ivs),R1*Q*R1',options_.qz_criterium);
% Pinf = bayestopt_.Pinf;
% by M. Ratto
RR=T(:,bayestopt_.restrict_var_list_nonstationary);
@ -245,15 +242,20 @@ function [fval,cost_flag,ys,trend_coeff,info] = DsgeLikelihood(xparam1,gend,data
LIK = DiffuseLikelihoodH3(T,R,Q,H,Pinf,Pstar,data,trend,start);
else
LIK = DiffuseLikelihoodH3corr(T,R,Q,H,Pinf,Pstar,data,trend,start);
end
end
end
end
else
if options_.kalman_algo == 1
%nv = size(bayestopt_.Z,1);
%LIK = kalman_filter(bayestopt_.Z,zeros(nv,nv),T,R,Q,data,zeros(size(T,1),1),Pstar,'u');
LIK = DiffuseLikelihood1(T,R,Q,Pinf,Pstar,data,trend,start);
% LIK = diffuse_likelihood1(T,R,Q,Pinf,Pstar,data-trend,start);
%if abs(LIK1-LIK)>0.0000000001
%tic
LIK = DiffuseLikelihood1(T,R,Q,Pinf,Pstar,data,trend,start);
%toc
%tic
%LIK1 = Diffuse_Likelihood1(T,R,Q,Pinf,Pstar,data,trend,start);
%toc
%LIK1-LIK
%if abs(LIK1-LIK)>0.0000000001
% disp(['LIK1 and LIK are not equal! ' num2str(abs(LIK1-LIK))])
%end
if isinf(LIK)

4
matlab/DsgeLikelihood_hh.m
View File

@ -160,7 +160,7 @@ function [fval,llik,cost_flag,ys,trend_coeff,info] = DsgeLikelihood(xparam1,gend
% 3. Initial condition of the Kalman filter
%------------------------------------------------------------------------------
if options_.lik_init == 1 % Kalman filter
Pstar = lyapunov_symm(T,R*Q*R');
Pstar = lyapunov_symm(T,R*Q*R',options_.qz_criterium);
Pinf = [];
elseif options_.lik_init == 2 % Old Diffuse Kalman filter
Pstar = 10*eye(np);
@ -183,7 +183,7 @@ function [fval,llik,cost_flag,ys,trend_coeff,info] = DsgeLikelihood(xparam1,gend
R1(j,:) = R1(j,:)-RR(j,i).*R(ivd(i),:);
end
end
Pstar = lyapunov_symm(T0,R1*Q*R1');
Pstar = lyapunov_symm(T0,R1*Q*R1',options_.qz_criterium);
end
%------------------------------------------------------------------------------
% 4. Likelihood evaluation

6
matlab/DsgeSmoother.m
View File

@ -1,6 +1,4 @@
function [alphahat,etahat,epsilonhat,ahat,SteadyState,trend_coeff,aK,T,R,P,PK,d,decomp] = DsgeSmoother(xparam1,gend,Y)
% function [alphahat,etahat,epsilonhat,ahat,SteadyState,trend_coeff,aK,T,R,P,PK,d,decomp] = DsgeSmoother(xparam1,gend,Y)
% Estimation of the smoothed variables and innovations.
%
% INPUTS
@ -95,7 +93,7 @@ function [alphahat,etahat,epsilonhat,ahat,SteadyState,trend_coeff,aK,T,R,P,PK,d,
H = M_.H;
if options_.lik_init == 1 % Kalman filter
Pstar = lyapunov_symm(T,R*Q*transpose(R));
Pstar = lyapunov_symm(T,R*Q*transpose(R),options_.qz_criterium);
Pinf = [];
elseif options_.lik_init == 2 % Old Diffuse Kalman filter
Pstar = 10*eye(np);
@ -105,7 +103,7 @@ function [alphahat,etahat,epsilonhat,ahat,SteadyState,trend_coeff,aK,T,R,P,PK,d,
Pstar = zeros(np,np);
ivs = bayestopt_.restrict_var_list_stationary;
R1 = R(ivs,:);
Pstar(ivs,ivs) = lyapunov_symm(T(ivs,ivs),R1*Q*R1');
Pstar(ivs,ivs) = lyapunov_symm(T(ivs,ivs),R1*Q*R1',options_.qz_criterium);
% Pinf = bayestopt_.Pinf;
% by M. Ratto
RR=T(:,bayestopt_.restrict_var_list_nonstationary);

2
matlab/DsgeVarLikelihood.m
View File

@ -122,7 +122,7 @@ end
%------------------------------------------------------------------------------
% 3. theoretical moments (second order)
%------------------------------------------------------------------------------
tmp0 = lyapunov_symm(T,R*Q*R');% I compute the variance-covariance matrix
tmp0 = lyapunov_symm(T,R*Q*R',options_.qz_criterium);% I compute the variance-covariance matrix
mf = bayestopt_.mf1; % of the restricted state vector.
% Get the non centered second order moments

91
matlab/UnivariateSpectralDensity.m
View File

@ -7,6 +7,7 @@ function [omega,f] = UnivariateSpectralDensity(dr,var_list)
% Adapted from th_autocovariances.m.
global options_ oo_ M_
omega = []; f = [];
if options_.order > 1
@ -17,8 +18,8 @@ if options_.order > 1
end
pltinfo = 1;%options_.SpectralDensity.Th.plot;
cutoff = 100;%options_.SpectralDensity.Th.cutoff;
sdl = 0.1;%options_.SepctralDensity.Th.sdl;
cutoff = 150;%options_.SpectralDensity.Th.cutoff;
sdl = 0.01;%options_.SepctralDensity.Th.sdl;
omega = (0:sdl:pi)';
GridSize = length(omega);
exo_names_orig_ord = M_.exo_names_orig_ord;
@ -76,34 +77,70 @@ for i=M_.maximum_lag:-1:2
AS(:,j1) = AS(:,j1)+ghx(:,i1);
i0 = i1;
end
b = ghu1*M_.Sigma_e*ghu1';
[A,B] = kalman_transition_matrix(dr);
% index of predetermined variables in A
i_pred = [nstatic+(1:npred) M_.endo_nbr+1:length(A)];
A = A(i_pred,i_pred);
[vx, ns_var] = lyapunov_symm(A,b);
i_ivar = find(~ismember(ivar,dr.order_var(ns_var+nstatic)));
ivar = ivar(i_ivar);
iky = iv(ivar);
Gamma = zeros(nvar,cutoff+1);
[A,B] = kalman_transition_matrix(dr,ikx',1:nx,dr.transition_auxiliary_variables,M_.exo_nbr);
[vx, u] = lyapunov_symm(A,B*M_.Sigma_e*B',options_.qz_criterium);
iky = iv(ivar);
if ~isempty(u)
iky = iky(find(any(abs(ghx(iky,:)*u) < options_.Schur_vec_tol,2)));
ivar = dr.order_var(iky);
end
aa = ghx(iky,:);
bb = ghu(iky,:);
Gamma = zeros(nvar,cutoff+1);
tmp = aa*vx*aa'+ bb*M_.Sigma_e*bb';
k = find(abs(tmp) < 1e-12);
tmp(k) = 0;
Gamma(:,1) = diag(tmp);
vxy = (A*vx*aa'+ghu1*M_.Sigma_e*bb');
tmp = aa*vxy;
k = find(abs(tmp) < 1e-12);
tmp(k) = 0;
Gamma(:,2) = diag(tmp);
for i=2:cutoff
vxy = A*vxy;
tmp = aa*vxy;
k = find(abs(tmp) < 1e-12);
tmp(k) = 0;
Gamma(:,i+1) = diag(tmp);
if options_.hp_filter == 0
tmp = aa*vx*aa'+ bb*M_.Sigma_e*bb';
k = find(abs(tmp) < 1e-12);
tmp(k) = 0;
Gamma(:,1) = diag(tmp);
vxy = (A*vx*aa'+ghu1*M_.Sigma_e*bb');
tmp = aa*vxy;
k = find(abs(tmp) < 1e-12);
tmp(k) = 0;
Gamma(:,2) = diag(tmp);
for i=2:cutoff
vxy = A*vxy;
tmp = aa*vxy;
k = find(abs(tmp) < 1e-12);
tmp(k) = 0;
Gamma(:,i+1) = diag(tmp);
end
else
iky = iv(ivar);
aa = ghx(iky,:);
bb = ghu(iky,:);
lambda = options_.hp_filter;
ngrid = options_.hp_ngrid;
freqs = 0 : ((2*pi)/ngrid) : (2*pi*(1 - .5/ngrid));
tpos = exp( sqrt(-1)*freqs);
tneg = exp(-sqrt(-1)*freqs);
hp1 = 4*lambda*(1 - cos(freqs)).^2 ./ (1 + 4*lambda*(1 - cos(freqs)).^2);
mathp_col = [];
IA = eye(size(A,1));
IE = eye(M_.exo_nbr);
for ig = 1:ngrid
f_omega =(1/(2*pi))*( [inv(IA-A*tneg(ig))*ghu1;IE]...
*M_.Sigma_e*[ghu1'*inv(IA-A'*tpos(ig)) IE]); % state variables
g_omega = [aa*tneg(ig) bb]*f_omega*[aa'*tpos(ig); bb']; % selected variables
f_hp = hp1(ig)^2*g_omega; % spectral density of selected filtered series
mathp_col = [mathp_col ; (f_hp(:))']; % store as matrix row
% for ifft
end;
imathp_col = real(ifft(mathp_col))*(2*pi);
tmp = reshape(imathp_col(1,:),nvar,nvar);
k = find(abs(tmp)<1e-12);
tmp(k) = 0;
Gamma(:,1) = diag(tmp);
sy = sqrt(Gamma(:,1));
sy = sy *sy';
for i=1:cutoff-1
tmp = reshape(imathp_col(i+1,:),nvar,nvar)./sy;
k = find(abs(tmp) < 1e-12);
tmp(k) = 0;
Gamma(:,i+1) = diag(tmp);
end
end
H = 1:cutoff;
for i=1:nvar
f(i,:) = Gamma(i,1)/(2*pi) + Gamma(i,H+1)*cos(H'*omega')/pi;

6
matlab/calib_obj.m
View File

@ -1,7 +1,7 @@
% targets and iy order: 1) variances 2) correlations
% 3) constraints on M_.Sigma_e itself 4) autocorrelations
function f=calib_obj(M_.Sigma_e,A,ghu1,ghx,ghu,targets,var_weights,iy,nar)
global vx fold
global vx fold options_
oo_.gamma_y = cell(nar+1,1);
% M_.Sigma_e = M_.Sigma_e'*M_.Sigma_e;
@ -10,11 +10,11 @@ function f=calib_obj(M_.Sigma_e,A,ghu1,ghx,ghu,targets,var_weights,iy,nar)
b=ghu1*M_.Sigma_e*ghu1';
vx = [];
if isempty(vx)
vx = lyapunov_symm(A,b);
vx = lyapunov_symm(A,b,options_.qz_criterium);
else
[vx,status] = bicgstab_(@f_var,b(:),vx(:),1e-8,50,A,nx);
if status
vx = lyapunov_symm(A,b);
vx = lyapunov_symm(A,b,options_.qz_criterium);
else
vx=reshape(vx,nx,nx);
end

4
matlab/calib_obj2.m
View File

@ -1,14 +1,14 @@
% targets and iy order: 1) variances 2) correlations
% 3) constraints on M_.Sigma_e itself 4) autocorrelations
function objective=calib_obj2(M_.Sigma_e,A,ghu1,ghx,ghu,targets,var_weights,iy,nar)
global vx fold
global vx fold options_
objective = cell (nar+3);
oo_.gamma_y = cell(nar+1,1);
M_.Sigma_e=diag(M_.Sigma_e);
nx = size(ghx,2);
b=ghu1*M_.Sigma_e*ghu1';
vx = lyapunov_symm(A,b);
vx = lyapunov_symm(A,b,options_.qz_criterium);
oo_.gamma_y{1} = ghx*vx*ghx'+ ghu*M_.Sigma_e*ghu';
if ~isempty(targets{1})
objective{1} = sqrt(oo_.gamma_y{1}(iy{1}));

80
matlab/check_posterior_analysis_data.m Normal file
View File

@ -0,0 +1,80 @@
function [info,description] = check_posterior_analysis_data(type,M_)
% part of DYNARE, copyright Dynare Team (2008)
% Gnu Public License.
info = 0;
if nargout>1
description = '';
end
%% Get informations about mcmc files.
if ~exist([ M_.dname '/metropolis'],'dir')
disp('check_posterior_analysis_data:: Can''t find any mcmc file!')
return
end
mhname = get_name_of_the_last_mh_file(M_);
mhdate = get_date_of_a_file(mhname);
%% Get informations about _posterior_draws files.
if ~exist([ M_.dname '/metropolis/' M_.fname '_posterior_draws.mat'])
info = 1; % select_posterior_draws has to be called first.
if nargout>1
description = 'select_posterior_draws has to be called.';
end
return
else
pddate = get_date_of_a_file([ M_.dname '/metropolis/' M_.fname '_posterior_draws.mat']);
if pddate<mhdate
info = 2; % _posterior_draws files have to be updated.
if nargout>1
description = 'posterior draws files have to be updated.';
end
return
else
info = 3; % Ok!
if nargout>1
description = 'posterior draws files are up to date.';
end
end
end
%% Get informations about posterior data files.
switch type
case 'variance'
generic_post_data_file_name = 'Posterior2ndOrderMoments';
case 'decomposition'
generic_post_data_file_name = 'PosteriorVarianceDecomposition';
case 'dynamic_decomposition'
generic_post_data_file_name = 'PosteriorDynamicVarianceDecomposition';
otherwise
disp('This feature is not yest implemented!')
end
pdfinfo = dir([ M_.dname '/metropolis/' M_.fname '_' generic_post_data_file_name '*'])
if isempty(pdfinfo)
info = 4; % posterior draws have to be processed.
if nargout>1
description = 'posterior draws files have to be processed.';
end
return
else
number_of_the_last_post_data_file = length(pdfinfo);
name_of_the_last_post_data_file = ...
[ M_.dname ...
'/metropolis/' ...
M_.dname ...
generic_post_data_file_name ...
int2str(number_of_the_last_post_data_file) ...
'.mat' ];
pdfdate = get_date_of_a_file(name_of_the_last_post_data_file);
if pdfdate<pddate
info = 5; % posterior data files have to be updated.
if nargout>1
description = 'posterior data files have to be updated.';
end
else
info = 6; % Ok (nothing to do ;-)
if nargout>1
description = 'There is nothing to do';
end
end
end

13
matlab/compute_model_moments.m
View File

@ -1,11 +1,9 @@
function moments=compute_model_moments(dr,options)
% function compute_model_moments(options)
% Computes posterior filter smoother and forecasts
function moments=compute_model_moments(dr,M_,options,_)
%
% INPUTS
% dr: structure describing model solution
% options: structure of Dynare options
% M_: structure of Dynare options
% options_
%
% OUTPUTS
% moments: a cell array containing requested moments
@ -16,7 +14,4 @@ function moments=compute_model_moments(dr,options)
% part of DYNARE, copyright Dynare Team (2008)
% Gnu Public License.
% subset of variables
varlist = options.varlist;
moments = th_autocovariances(dr,varlist);
moments = th_autocovariances(dr,options_.varlist,M_,options_);

2
matlab/disp_th_moments.m
View File

@ -19,7 +19,7 @@ function disp_th_moments(dr,var_list)
end
end
[oo_.gamma_y,ivar] = th_autocovariances(dr,ivar);
[oo_.gamma_y,ivar] = th_autocovariances(dr,ivar,M_,options_);
m = dr.ys(ivar);

97
matlab/dsge_posterior_theoretical_covariance.m
View File

@ -1,10 +1,12 @@
function dsge_posterior_theoretical_covariance()
function [nvar,vartan,CovarFileNumber] = dsge_posterior_theoretical_covariance(SampleSize,M_,options_,oo_)
% This function estimates the posterior density of the endogenous
% variables second order moments.
%
% INPUTS
% None.
%
% SampleSize [integer]
%
%
%
% OUTPUTS
% None.
%
@ -21,52 +23,30 @@ function dsge_posterior_theoretical_covariance()
% part of DYNARE, copyright Dynare Team (2007-2008)
% Gnu Public License.
global M_ options_ oo_
type = 'posterior';% To be defined as a input argument later...
NumberOfSimulations = 1000;% To be defined in a global structure...
type = 'posterior';
% Set varlist (vartan)
[ivar,vartan] = set_stationary_variables_list;
% Set various parameters & Check or create files and directories
if strcmpi(type,'posterior')
MhDirectoryName = CheckPath('metropolis');
else
MhDirectoryName = CheckPath('prior');
end
fname = [ MhDirectoryName '/' M_.fname];
%save([fname '_Posterior2ndOrder'],'varlist');
DrawsFiles = dir([fname '_' type '_draws*' ]);
if ~rows(DrawsFiles)
if strcmpi(type,'posterior')
drsize = size_of_the_reduced_form_model(oo_.dr);
if drsize*NumberOfSimulations>101 %Too big model!
drsize=0;
end
SampleAddress = selec_posterior_draws(NumberOfSimulations,drsize);
else
% (samples from the prior) To be done later...
end
DrawsFiles = dir([fname '_' type '_draws*']);
end
% Get the number of stationary endogenous variables.
nvar = length(ivar);
nar = options_.ar;% Saves size of the auto-correlation function.
options_.ar = 0;% Set the size of the auto-correlation function to zero.
% Set the size of the auto-correlation function to zero.
nar = options_.ar;
options_.ar = 0;
NumberOfDrawsFiles = rows(DrawsFiles);
% Get informations about the _posterior_draws files.
DrawsFiles = dir([M_.dname '/metropolis/' M_.fname '_' type '_draws*' ]);
NumberOfDrawsFiles = length(DrawsFiles);
% Number of lines in posterior data files.
MaXNumberOfCovarLines = ceil(options_.MaxNumberOfBytes/(nvar*(nvar+1)/2)/8);
if NumberOfSimulations<=MaXNumberOfCovarLines
if SampleSize<=MaXNumberOfCovarLines
Covariance_matrix = zeros(NumberOfSimulations,nvar*(nvar+1)/2);
NumberOfCovarFiles = 1;
else
Covariance_matrix = zeros(MaXNumberOfCovarLines,nvar*(nvar+1)/2);
NumberOfLinesInTheLastCovarFile = mod(NumberOfSimulations,MaXNumberOfCovarLines);
NumberOfCovarFiles = ceil(NumberOfSimulations/MaXNumberOfCovarLines);
NumberOfLinesInTheLastCovarFile = mod(SampleSize,MaXNumberOfCovarLines);
NumberOfCovarFiles = ceil(SampleSize/MaXNumberOfCovarLines);
end
NumberOfCovarLines = rows(Covariance_matrix);
@ -86,10 +66,10 @@ for file = 1:NumberOfDrawsFiles
set_parameters(pdraws{linee,1});
[dr,info] = resol(oo_.steady_state,0);
end
tmp = th_autocovariances(dr,ivar);
tmp = th_autocovariances(dr,ivar,M_,options_);
for i=1:nvar
for j=i:nvar
Covariance_matrix(linea,idx(i,j,nvar)) = tmp{1}(i,j);
Covariance_matrix(linea,symmetric_matrix_index(i,j,nvar)) = tmp{1}(i,j);
end
end
if linea == NumberOfCovarLines
@ -109,45 +89,12 @@ for file = 1:NumberOfDrawsFiles
end
end
end
options_.ar = nar; clear('pdraws','tmp');
% Compute statistics and save in oo_
for i=1:nvar
for j=i:nvar
i1 = 1;
tmp = zeros(NumberOfSimulations,1);
for file = 1:CovarFileNumber
load([fname '_Posterior2ndOrderMoments' int2str(file)]);
i2 = i1 + rows(Covariance_matrix) - 1;
tmp(i1:i2) = Covariance_matrix(:,idx(i,j,nvar));
i1 = i2+1;
end
[post_mean, post_median, post_var, hpd_interval, post_deciles, density] = ...
posterior_moments(tmp,1,options_.mh_conf_sig);
name = fieldname(i,j,vartan);
eval(['oo_.PosteriorTheoreticalMoments.dsge.covariance.mean.' name ' = post_mean;']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.covariance.median.' name ' = post_median;']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.covariance.variance.' name ' = post_var;']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.covariance.hpdinf.' name ' = hpd_interval(1);']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.covariance.hpdsup.' name ' = hpd_interval(2);']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.covariance.deciles.' name ' = post_deciles;']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.covariance.density.' name ' = density;']);
end
end
options_.ar = nar;
function k = idx(i,j,n)
k = (i-1)*n+j-i*(i-1)/2;
function r = rows(M)
r = size(M,1);
function c = cols(M)
c = size(M,2);
function name = fieldname(i,j,vlist)
n1 = deblank(vlist(i,:));
n2 = deblank(vlist(j,:));
name = [n1 '.' n2];
c = size(M,2);

103
matlab/dsge_posterior_theoretical_variance_decomposition.m
View File

@ -1,4 +1,5 @@
function dsge_posterior_theoretical_variance_decomposition()
function [nvar,vartan,NumberOfDecompFiles] = ...
dsge_posterior_theoretical_variance_decomposition(SampleSize,M_,options_,oo_)
% This function estimates the posterior distribution of the variance
% decomposition of the observed endogenous variables.
%
@ -20,42 +21,22 @@ function dsge_posterior_theoretical_variance_decomposition()
% part of DYNARE, copyright Dynare Team (2007-2008).
% Gnu Public License.
global M_ options_ oo_
type = 'posterior';% To be defined as a input argument later...
NumberOfSimulations = 800;% To be defined in a global structure...
% Set varlist (vartan)
[ivar,vartan] = set_stationary_variables_list;
ivar
vartan
nvar = length(ivar);
% Set various parameters, Check or create files and directories &
% initialize arrays.
if strcmpi(type,'posterior')
MhDirectoryName = CheckPath('metropolis');
else
MhDirectoryName = CheckPath('prior');
end
fname = [ MhDirectoryName '/' M_.fname];
DrawsFiles = dir([fname '_' type '_draws*' ]);
if ~rows(DrawsFiles)
if strcmpi(type,'posterior')
drsize = size_of_the_reduced_form_model(oo_.dr);
if drsize*NumberOfSimulations>101%Big model!
drsize=0;
end
SampleAddress = selec_posterior_draws(NumberOfSimulations,drsize);
else% (samples from the prior) To be done later...
end
DrawsFiles = dir([fname '_' type '_draws*']);
end
% Set the size of the auto-correlation function to zero.
nar = options_.ar;
options_.ar = 0;
nar = options_.ar;% Saves size of the auto-correlation function.
options_.ar = 0;% Set the size of the auto-correlation function.
% Get informations about the _posterior_draws files.
DrawsFiles = dir([M_.dname '/metropolis/' M_.fname '_' type '_draws*' ]);
NumberOfDrawsFiles = length(DrawsFiles);
nexo = M_.exo_nbr;
nvar = length(ivar);
NumberOfDrawsFiles = rows(DrawsFiles);
NumberOfSavedElementsPerSimulation = nvar*(nexo+1);
@ -89,10 +70,11 @@ for file = 1:NumberOfDrawsFiles
set_parameters(pdraws{linee,1});
[dr,info] = resol(oo_.steady_state,0);
end
tmp = th_autocovariances(dr,ivar);
for i=1:nvar
Decomposition_array(linea,i) = tmp{1}(i,i);
end
tmp = th_autocovariances(dr,ivar,M_,options_);
%for i=1:nvar
% Decomposition_array(linea,i) = tmp{1}(i,i);
%end
Decomposition_array(linea,:) = transpose(tmp{1});
for i=1:nvar
for j=1:nexo
Decomposition_array(linea,nvar+(i-1)*nexo+j) = tmp{2}(i,j);
@ -115,62 +97,11 @@ for file = 1:NumberOfDrawsFiles
end
end
end
options_.ar = nar; clear('pdraws','tmp');
% Compute statistics and save in oo_
for i=1:nvar
for j=1:nexo
i1 = 1;
tmp = zeros(NumberOfSimulations,1);
for file = 1:DecompFileNumber
load([fname '_PosteriorVarianceDecomposition' int2str(file)]);
i2 = i1 + rows(Decomposition_array) - 1;
tmp(i1:i2) = Decomposition_array(:,nvar+(i-1)*nexo+j);
i1 = i2+1;
end
name = [ deblank(vartan(i,:)) '.' deblank(M_.exo_names(j,:)) ];
t1 = min(tmp); t2 = max(tmp);
t3 = t2-t1;% How to normalize ? t1 and t2 may be zero...
if t3<1.0e-12
if t1<1.0e-12
t1 = 0;
end
if abs(t1-1)<1.0e-12
t1 = 1;
end
post_mean = t1;
post_median = t1;
post_var = 0;
hpd_interval = NaN(2,1);
post_deciles = NaN(9,1);
density = NaN;
else
[post_mean, post_median, post_var, hpd_interval, post_deciles, density] = ...
posterior_moments(tmp,1,options_.mh_conf_sig);
end
eval(['oo_.PosteriorTheoreticalMoments.dsge.VarianceDecomposition.mean.' name ' = post_mean;']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.VarianceDecomposition.median.' name ' = post_median;']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.VarianceDecomposition.variance.' name ' = post_var;']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.VarianceDecomposition.hpdinf.' name ' = hpd_interval(1);']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.VarianceDecomposition.hpdsup.' name ' = hpd_interval(2);']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.VarianceDecomposition.deciles.' name ' = post_deciles;']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.VarianceDecomposition.density.' name ' = density;']);
end
end
function k = idx(i,j,n)
k = (i-1)*n+j-i*(i-1)/2;
options_.ar = nar;% Useless because options_ is not a global anymore...
function r = rows(M)
r = size(M,1);
function c = cols(M)
c = size(M,2);
function name = fieldname(i,j,vlist)
n1 = deblank(vlist(i,:));
n2 = deblank(vlist(j,:));
name = [n1 '.' n2];
c = size(M,2);

4
matlab/dynare_resolve.m
View File

@ -32,7 +32,7 @@ function [A,B,ys,info] = dynare_resolve(iv,ic,aux)
global oo_ M_
[oo_.dr,info] = resol(oo_.steady_state,0);
if info(1) > 0
A = [];
B = [];
@ -52,5 +52,5 @@ function [A,B,ys,info] = dynare_resolve(iv,ic,aux)
aux(k,2) = aux(k,2) + oo_.dr.nfwrd;
end
[A,B] = kalman_transition_matrix(oo_.dr,iv,ic,aux);
[A,B] = kalman_transition_matrix(oo_.dr,iv,ic,aux,M_.exo_nbr);
ys = oo_.dr.ys;

2
matlab/forcst.m
View File

@ -30,7 +30,7 @@ function [yf,int_width]=forcst(dr,y0,horizon,var_list)
nc = size(dr.ghx,2);
endo_nbr = M_.endo_nbr;
inv_order_var = dr.inv_order_var;
[A,B] = kalman_transition_matrix(dr,nstatic+(1:npred),1:nc,dr.transition_auxiliary_variables);
[A,B] = kalman_transition_matrix(dr,nstatic+(1:npred),1:nc,dr.transition_auxiliary_variables,M_.exo_nbr);
nvar = size(var_list,1);
if nvar == 0

8
matlab/get_date_of_a_file.m Normal file
View File

@ -0,0 +1,8 @@
function [d1,d2] = get_date_of_a_file(filename)
% part of DYNARE, copyright Dynare Team (2008)
% Gnu Public License.
info = dir(filename);
d1 = info.datenum;
if nargout>1
d2 = info.date;
end

17
matlab/get_innovation_contemporaneous_impact.m
View File

@ -1,4 +1,4 @@
function F = get_innovation_contemporaneous_impact('type')
function B = get_innovation_contemporaneous_impact(type,info)
% function F = get_innovation_contemporaneous_impact('type')
% The approximated reduced form model is
@ -15,7 +15,8 @@ function F = get_innovation_contemporaneous_impact('type')
% given by F = Z*B. Matrix F is returned by this function.
%
% INPUTS
% o type = "mode","mean"
% o type [string] "mode","mean"
% o info [integer] if equal to 1, matrix B is saved in a mat file.
%
% OUTPUTS
% o F (F is also saved in a file)
@ -26,17 +27,21 @@ function F = get_innovation_contemporaneous_impact('type')
% part of DYNARE, copyright Dynare Team (2006-2008)
% Gnu Public License.
global oo_ M_ bayestopt_
global oo_ M_ bayestopt_ options_
if nargin == 0
type = 'mode';
end
if nargin == 1
info = 0;
end
get_posterior_parameters(type);
[dr,info]=dr1(oo_.dr,0);
[dr,info,M_,options_,oo_]=dr1(oo_.dr,0,M_,options_,oo_);
B(dr.order_var,M_.exo_names_orig_ord) = dr.ghu*sqrt(M_.Sigma_e);
F = B(bayestopt_.mfys,:);
B = B(bayestopt_.mfys,:);
save([M_.fname '_InnovImpact',F]);
save([M_.fname '_InnovImpact'],'B');

1
matlab/get_moments_size.m
View File

@ -1,5 +1,4 @@
function s=get_moments_size(options)
% function PosteriorFilterSmootherAndForecast(Y,gend, type)
% Computes posterior filter smoother and forecasts
%

17
matlab/get_name_of_the_last_mh_file.m Normal file
View File

@ -0,0 +1,17 @@
function mhname = get_name_of_the_last_mh_file(M_)
% part of DYNARE, copyright Dynare Team (2008)
% Gnu Public License.
model_name = M_.fname ;
mcmc_directory = M_.dname ;
load([ mcmc_directory '/metropolis/' model_name '_mh_history']) ;
mh_number = record.LastFileNumber ;
bk_number = record.Nblck ;
clear('record') ;
mhname = [ mcmc_directory ...
'/metropolis/' ...
model_name ...
'_mh' ...
int2str(mh_number) ...
'_blck' ...
int2str(bk_number) ...
'.mat'] ;

4
matlab/get_variance_of_endogenous_variables.m
View File

@ -29,9 +29,9 @@ function [vx1,i_ns] = get_variance_of_endogenous_variables(dr,i_var)
nc = size(ghx,2);
n = length(i_var);
[A,B] = kalman_transition_matrix(dr,nstatic+(1:npred),1:nc,dr.transition_auxiliary_variables);
[A,B] = kalman_transition_matrix(dr,nstatic+(1:npred),1:nc,dr.transition_auxiliary_variables,M_.exo_nbr);
[vx,u] = lyapunov_symm(A,B*Sigma_e*B');
[vx,u] = lyapunov_symm(A,B*Sigma_e*B',options_.qz_criterium);
if size(u,2) > 0
i_stat = find(any(abs(ghx*u) < options_.Schur_vec_tol,2));

2
matlab/global_initialization.m
View File

@ -145,6 +145,8 @@ function global_initialization()
options_.cutoff = 1e-12;
options_.student_degrees_of_freedom = 3;
options_.subdraws = [];
options_.PosteriorSampleSize = 1000;
options_.MaximumNumberOfMegaBytes = 111;
% Misc
options_.conf_sig = 0.6;

13
matlab/kalman_transition_matrix.m
View File

@ -1,14 +1,13 @@
function [A,B] = kalman_transition_matrix(dr,iv,ic,aux)
% function [A,B] = kalman_transition_matrix(dr,iv,ic,aux)
% makes transition matrices out of ghx and ghu for Kalman filter
function [A,B] = kalman_transition_matrix(dr,iv,ic,aux,exo_nbr)
% Builds the transition equation of the state space representation out of ghx and ghu for Kalman filter
%
% INPUTS
% dr: structure of decisions rules for stochastic simulations
% iv: selected variables
% ic: state variables position in the transition matrix columns
% aux: indices for auxiliary equations
%
% exo_nbr: number of exogenous variables
%
% OUTPUTS
% A: matrix of predetermined variables effects in linear solution (ghx)
% B: matrix of shocks effects in linear solution (ghu)
@ -18,15 +17,13 @@ function [A,B] = kalman_transition_matrix(dr,iv,ic,aux)
%
% part of DYNARE, copyright Dynare Team (2003-2008)
% Gnu Public License.
global M_
n_iv = length(iv);
n_ir1 = size(aux,1);
nr = n_iv + n_ir1;
A = zeros(nr,nr);
B = zeros(nr,M_.exo_nbr);
B = zeros(nr,exo_nbr);
i_n_iv = 1:n_iv;
A(i_n_iv,ic) = dr.ghx(iv,:);

14
matlab/lyapunov_symm.m
View File

@ -1,7 +1,5 @@
function [x,u]=lyapunov_symm(a,b)
% function [x,u]=lyapunov_symm(a,b)
% solves the Lyapunov equation x-a*x*a' = b, for b (and then x) symmetrical
function [x,u]=lyapunov_symm(a,b,qz_criterium)
% Solves the Lyapunov equation x-a*x*a' = b, for b (and then x) symmetrical
% if a has some unit roots, the function computes only the solution of the stable subsystem
%
% INPUTS:
@ -21,9 +19,7 @@ function [x,u]=lyapunov_symm(a,b)
% needs Matlab version with ordeig function
%
% part of DYNARE, copyright Dynare Team (2006-2008)
% Gnu Public License
global options_
% Gnu Public License
ns_var = [];
u = [];
@ -35,9 +31,9 @@ function [x,u]=lyapunov_symm(a,b)
end
[u,t] = schur(a);
if exist('ordeig','builtin')
e1 = abs(ordeig(t)) > 2-options_.qz_criterium;
e1 = abs(ordeig(t)) > 2-qz_criterium;
else
e1 = abs(my_ordeig(t)) > 2-options_.qz_criterium;
e1 = abs(my_ordeig(t)) > 2-qz_criterium;
end
k = sum(e1);
if exist('ordschur','builtin')

137
matlab/posterior_analysis.m Normal file
View File

@ -0,0 +1,137 @@
function posterior_analysis(type,arg1,arg2,options_,M_,oo_)
% part of DYNARE, copyright Dynare Team (2008)
% Gnu Public License.
info = check_posterior_analysis_data(type,M_);
switch info
case 0
disp('check_posterior_analysis_data:: Can''t find any mcmc file!')
error('Check the options of the estimation command...')
case {1,2}
SampleSize = options_.PosteriorSampleSize;
MaxMegaBytes = options_.MaximumNumberOfMegaBytes;
drsize = size_of_the_reduced_form_model(oo_.dr);
if drsize*SampleSize>MaxMegaBytes
drsize=0;
end
SampleAddress = selec_posterior_draws(SampleSize,drsize);
case {4,5}
switch type
case 'variance'
[nvar,vartan,NumberOfFiles] = ...
dsge_posterior_theoretical_covariance(SampleSize,M_,options_,oo_);
case 'decomposition'
[nvar,vartan,NumberOfFiles] = ...
dsge_posterior_theoretical_variance_decomposition(SampleSize,M_,options_,oo_);
otherwise
disp('Not yet implemented')
end
case 6
switch type
case 'variance'
covariance_posterior_analysis(NumberOfFiles,SampleSize,M_.dname,M_.fname,...
vartan,nvar,arg1,arg2);
case 'decomposition'
variance_decomposition_posterior_analysis(NumberOfFiles,SampleSize,M_.dname,M_.fname,...
M_.exo_names,arg2,vartan,nvar,arg1);
otherwise
disp('Not yet implemented')
end
otherwise
error(['posterior_analysis:: Check_posterior_analysis_data gave a meaningless output!'])
end
function covariance_posterior_analysis(NumberOfFiles,NumberOfSimulations,dname,fname,vartan,nvar,var1,var2)
indx1 = check_name(vartan,var1)
if isempty(indx1)
disp(['posterior_analysis:: ' var1 ' is not a stationary endogenous variable!'])
return
end
if ~isempty(var2)
indx2 = check_name(vartan,var2)
if isempty(indx2)
disp(['posterior_analysis:: ' var2 ' is not a stationary endogenous variable!'])
return
end
else
indx2 = indx1;
var2 = var1;
end
i1 = 1; tmp = zeros(NumberOfSimulations,1);
for file = 1:CovarFileNumber
load([ dname '/metropolis/' fname '_Posterior2ndOrderMoments' int2str(file)]);
i2 = i1 + rows(Covariance_matrix) - 1;
tmp(i1:i2) = Covariance_matrix(:,symmetric_matrix_index(indx1,indx2,nvar));
i1 = i2+1;
end
[post_mean, post_median, post_var, hpd_interval, post_deciles, density] = ...
posterior_moments(tmp,1,options_.mh_conf_sig);
if strcmpi(var1,var2)
name = var1;
else
name = [var1 '.' var2];
end
eval(['oo_.PosteriorTheoreticalMoments.dsge.covariance.mean.' name ' = post_mean;']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.covariance.median.' name ' = post_median;']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.covariance.variance.' name ' = post_var;']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.covariance.hpdinf.' name ' = hpd_interval(1);']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.covariance.hpdsup.' name ' = hpd_interval(2);']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.covariance.deciles.' name ' = post_deciles;']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.covariance.density.' name ' = density;']);
function variance_decomposition_posterior_analysis(NumberOfFiles,NumberOfSimulations,dname,fname, ...
exonames,exo,vartan,nvar,var)
indx1 = check_name(vartan,var)
if isempty(indx)
disp(['posterior_analysis:: ' var ' is not a stationary endogenous variable!'])
return
end
jndx = check_name(exonames,exo);
if isempty(jndx)
disp(['posterior_analysis:: ' exo ' is not a declared exogenous variable!'])
return
end
i1 = 1; tmp = zeros(NumberOfSimulations,1);
for file = 1:NumberOfFiles
load([fname '_PosteriorVarianceDecomposition' int2str(file)]);
i2 = i1 + rows(Decomposition_array) - 1;
tmp(i1:i2) = Decomposition_array(:,nvar+(i-1)*nexo+j);
i1 = i2+1;
end
name = [ var '.' exo ];
t1 = min(tmp); t2 = max(tmp);
t3 = t2-t1;% How to normalize ? t1 and t2 may be zero...
if t3<1.0e-12
if t1<1.0e-12
t1 = 0;
end
if abs(t1-1)<1.0e-12
t1 = 1;
end
post_mean = t1;
post_median = t1;
post_var = 0;
hpd_interval = NaN(2,1);
post_deciles = NaN(9,1);
density = NaN;
else
[post_mean, post_median, post_var, hpd_interval, post_deciles, density] = ...
posterior_moments(tmp,1,options_.mh_conf_sig);
end
eval(['oo_.PosteriorTheoreticalMoments.dsge.VarianceDecomposition.mean.' name ' = post_mean;']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.VarianceDecomposition.median.' name ' = post_median;']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.VarianceDecomposition.variance.' name ' = post_var;']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.VarianceDecomposition.hpdinf.' name ' = hpd_interval(1);']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.VarianceDecomposition.hpdsup.' name ' = hpd_interval(2);']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.VarianceDecomposition.deciles.' name ' = post_deciles;']);
eval(['oo_.PosteriorTheoreticalMoments.dsge.VarianceDecomposition.density.' name ' = density;']);
function n = check_name(vartan,varname)
n = strmatch(varname,vartan,'exact')

3
matlab/prior_posterior_statistics.m
View File

@ -127,12 +127,11 @@ for b=1:B
set_all_parameters(deep);
dr = resol(oo_.steady_state,0);
if moments_varendo
stock_moments{irun(8)} = compute_model_moments(dr,options_);
stock_moments{irun(8)} = compute_model_moments(dr,M_,options_);
end
if run_smoother
[alphahat,etahat,epsilonhat,ahat,SteadyState,trend_coeff,aK] = ...
DsgeSmoother(deep,gend,Y);
if options_.loglinear
stock_smooth(dr.order_var,:,irun(1)) = alphahat(1:endo_nbr,:)+ ...
repmat(log(dr.ys(dr.order_var)),1,gend);

4
matlab/symmetric_matrix_index.m Normal file
View File

@ -0,0 +1,4 @@
function k = symmetric_matrix_index(i,j,n)
% part of DYNARE, copyright Dynare Team (2007-2008).
% Gnu Public License.
k = (i-1)*n+j-i*(i-1)/2;

25
matlab/th_autocovariances.m
View File

@ -1,14 +1,14 @@
function [Gamma_y,ivar]=th_autocovariances(dr,ivar)
% function [Gamma_y,ivar]=th_autocovariances(dr,ivar)
% computes the theoretical auto-covariances, Gamma_y, for an AR(p) process
function [Gamma_y,ivar]=th_autocovariances(dr,ivar,M_,options_)
% Computes the theoretical auto-covariances, Gamma_y, for an AR(p) process
% with coefficients dr.ghx and dr.ghu and shock variances Sigma_e_
% for a subset of variables ivar (indices in lgy_)
%
% INPUTS
% dr: structure of decisions rules for stochastic simulations
% ivar: subset of variables
%
% M_
% options_
%
% OUTPUTS
% Gamma_y: theoritical auto-covariances
% ivar: subset of variables
@ -19,9 +19,6 @@ function [Gamma_y,ivar]=th_autocovariances(dr,ivar)
% part of DYNARE, copyright Dynare Team (2001-2008)
% Gnu Public License.
global M_ options_
exo_names_orig_ord = M_.exo_names_orig_ord;
if sscanf(version('-release'),'%d') < 13
warning off
@ -68,9 +65,9 @@ function [Gamma_y,ivar]=th_autocovariances(dr,ivar)
ipred = nstatic+(1:npred)';
% state space representation for state variables only
[A,B] = kalman_transition_matrix(dr,ipred,1:nx,dr.transition_auxiliary_variables);
[A,B] = kalman_transition_matrix(dr,ipred,1:nx,dr.transition_auxiliary_variables,M_.exo_nbr);
if options_.order == 2 | options_.hp_filter == 0
[vx, u] = lyapunov_symm(A,B*M_.Sigma_e*B');
[vx, u] = lyapunov_symm(A,B*M_.Sigma_e*B',options_.qz_criterium);
iky = iv(ivar);
if ~isempty(u)
iky = iky(find(all(abs(ghx(iky,:)*u) < options_.Schur_vec_tol,2)));
@ -113,10 +110,10 @@ function [Gamma_y,ivar]=th_autocovariances(dr,ivar)
b1 = b1*cs;
b2(:,exo_names_orig_ord) = ghu(iky,:);
b2 = b2*cs;
vx = lyapunov_symm(A,b1*b1');
vx = lyapunov_symm(A,b1*b1',options_.qz_criterium);
vv = diag(aa*vx*aa'+b2*b2');
for i=1:M_.exo_nbr
vx1 = lyapunov_symm(A,b1(:,i)*b1(:,i)');
vx1 = lyapunov_symm(A,b1(:,i)*b1(:,i)',options_.qz_criterium);
Gamma_y{nar+2}(:,i) = abs(diag(aa*vx1*aa'+b2(:,i)*b2(:,i)'))./vv;
end
end
@ -161,9 +158,9 @@ function [Gamma_y,ivar]=th_autocovariances(dr,ivar)
end
%variance decomposition
if M_.exo_nbr > 1
if M_.exo_nbr > 1
Gamma_y{nar+2} = zeros(nvar,M_.exo_nbr);
SS(exo_names_orig_ord,exo_names_orig_ord)=M_.Sigma_e+1e-14*eye(M_.exo_nbr);
SS(exo_names_orig_ord,exo_names_orig_ord) = M_.Sigma_e+1e-14*eye(M_.exo_nbr);
cs = chol(SS)';
SS = cs*cs';
b1(:,exo_names_orig_ord) = ghu1;