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

Merge branch 'dynare-ident_globals' 

Ref. !2213
new-samplers
Sébastien Villemot 2023-12-07 12:19:04 +01:00
parent 6a58316a3e 1e5a04c2c5
commit 40ef192e37
No known key found for this signature in database
GPG Key ID: 2CECE9350ECEBE4A
79 changed files with 1435 additions and 3251 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
license.txt
View File

@ -171,17 +171,13 @@ Comment: Written by Jessica Cariboni and Francesca Campolongo
Files: matlab/gsa/cumplot.m
matlab/gsa/filt_mc_.m
matlab/gsa/gsa_plotmatrix.m
matlab/gsa/gsa_skewness.m
matlab/gsa/gsa_speed.m
matlab/gsa/log_trans_.m
matlab/gsa/map_calibration.m
matlab/gsa/map_ident_.m
matlab/gsa/mcf_analysis.m
matlab/gsa/myboxplot.m
matlab/gsa/myprctilecol.m
matlab/gsa/prior_draw_gsa.m
matlab/gsa/read_data.m
matlab/gsa/redform_map.m
matlab/gsa/redform_screen.m
matlab/gsa/scatter_mcf.m
@ -192,26 +188,10 @@ Files: matlab/gsa/cumplot.m
matlab/gsa/stand_.m
matlab/gsa/tcrit.m
matlab/gsa/teff.m
matlab/gsa/trank.m
Copyright: 2011-2018 European Commission
2011-2018 Dynare Team
2011-2023 Dynare Team
License: GPL-3+
Files: matlab/gsa/pick.m
Copyright: none
License: public-domain-jrc
This software has been developed at the Joint Research Centre of European Commission
by officers in the course of their official duties. This software is not subject to copyright
protection and is in the public domain. It is an experimental system. The Joint Research Centre
of European Commission assumes no responsibility whatsoever for its use by other parties
and makes no guarantees, expressed or implied, about its quality, reliability, or any other
characteristic. We would appreciate acknowledgement if the software is used.
Comment: This file is part of GLUEWIN.
The program has been developed by M. Ratto, European Commission, Joint Research Centre,
Institute for the Protection and Security of The Citizen, Technological and Economic Risk Management,
Applied Statistics, as a deliverable of the IMPACT project
(EC Fifth Framework Programme, SCA Project, IST-1999-11313, DG-INFSO).
Files: matlab/optimization/simpsa.m matlab/optimization/simpsaget.m matlab/optimization/simpsaset.m
Copyright: 2005 Henning Schmidt, FCC, henning@fcc.chalmers.se
2006 Brecht Donckels, BIOMATH, brecht.donckels@ugent.be

4
matlab/+mom/standard_errors.m
View File

@ -62,7 +62,7 @@ if strcmp(options_mom_.mom.mom_method,'GMM') && options_mom_.mom.analytic_standa
idx_nan = find(any(isnan(D)));
if any(idx_nan)
for i = idx_nan
fprintf('No standard errors available for parameter %s\n',get_the_name(i,options_mom_.TeX, M_, estim_params_, options_mom_))
fprintf('No standard errors available for parameter %s\n',get_the_name(i,options_mom_.TeX, M_, estim_params_, options_mom_.varobs))
end
warning('There are NaN in the analytical Jacobian of Moments. Check your bounds and/or priors, or use a different optimizer.')
Asympt_Var = NaN(length(xparam),length(xparam));
@ -86,7 +86,7 @@ else
if nnz(info_p)==0 && nnz(info_m)==0
D(:,i) = (oo__p.mom.model_moments - oo__m.mom.model_moments)/(2*eps_value);
else
problpar = get_the_name(i,options_mom_.TeX, M_, estim_params_, options_mom_);
problpar = get_the_name(i,options_mom_.TeX, M_, estim_params_, options_mom_.varobs);
if info_p(1)==42
warning('method_of_moments:info','Cannot compute the Jacobian using finite differences for parameter %s due to hitting the upper bound - no standard errors available.\n',problpar)
else

4
matlab/DSMH_sampler.m
View File

@ -98,7 +98,7 @@ TeX = options_.TeX;
str = sprintf(' Param. \t Lower Bound (95%%) \t Mean \t Upper Bound (95%%)');
for l=1:npar
[name,~] = get_the_name(l,TeX,M_,estim_params_,options_);
[name,~] = get_the_name(l,TeX,M_,estim_params_,options_.varobs);
str = sprintf('%s\n %s \t\t %5.4f \t\t %7.5f \t\t %5.4f', str, name, lb95_xparam(l), mean_xparam(l), ub95_xparam(l));
end
disp([str])
@ -129,7 +129,7 @@ hh_fig = dyn_figure(options_.nodisplay,'Name','Parameters Densities');
for k=1:npar %min(nstar,npar-(plt-1)*nstar)
subplot(ceil(sqrt(npar)),floor(sqrt(npar)),k)
%kk = (plt-1)*nstar+k;
[name,texname] = get_the_name(k,TeX,M_,estim_params_,options_);
[name,texname] = get_the_name(k,TeX,M_,estim_params_,options_.varobs);
optimal_bandwidth = mh_optimal_bandwidth(distrib_param(k,:)',options_.posterior_sampler_options.dsmh.nparticles,bandwidth,kernel_function);
[density(:,1),density(:,2)] = kernel_density_estimate(distrib_param(k,:)',number_of_grid_points,...
options_.posterior_sampler_options.dsmh.nparticles,optimal_bandwidth,kernel_function);

4
matlab/Herbst_Schorfheide_sampler.m
View File

@ -126,7 +126,7 @@ TeX = options_.TeX;
str = sprintf(' Param. \t Lower Bound (95%%) \t Mean \t Upper Bound (95%%)');
for l=1:npar
[name,~] = get_the_name(l,TeX,M_,estim_params_,options_);
[name,~] = get_the_name(l,TeX,M_,estim_params_,options_.varobs);
str = sprintf('%s\n %s \t\t %5.4f \t\t %7.5f \t\t %5.4f', str, name, lb95_xparam(l), mean_xparam(l), ub95_xparam(l));
end
disp([str])
@ -156,7 +156,7 @@ hh_fig = dyn_figure(options_.nodisplay,'Name','Parameters Densities');
for k=1:npar %min(nstar,npar-(plt-1)*nstar)
subplot(ceil(sqrt(npar)),floor(sqrt(npar)),k)
%kk = (plt-1)*nstar+k;
[name,texname] = get_the_name(k,TeX,M_,estim_params_,options_);
[name,texname] = get_the_name(k,TeX,M_,estim_params_,options_.varobs);
optimal_bandwidth = mh_optimal_bandwidth(distrib_param(k,:)',options_.posterior_sampler_options.HSsmc.nparticles,bandwidth,kernel_function);
[density(:,1),density(:,2)] = kernel_density_estimate(distrib_param(k,:)',number_of_grid_points,...
options_.posterior_sampler_options.HSsmc.nparticles,optimal_bandwidth,kernel_function);

8
matlab/PlotPosteriorDistributions.m
View File

@ -1,6 +1,5 @@
function oo_ = PlotPosteriorDistributions(estim_params_, M_, options_, bayestopt_, oo_)
% function PlotPosteriorDistributions()
% oo_ = PlotPosteriorDistributions(estim_params_, M_, options_, bayestopt_, oo_)
% plots posterior distributions
%
% INPUTS
@ -36,7 +35,6 @@ latexDirectoryName = CheckPath('latex',M_.dname);
graphDirectoryName = CheckPath('graphs',M_.dname);
TeX = options_.TeX;
nblck = options_.mh_nblck;
nvx = estim_params_.nvx;
nvn = estim_params_.nvn;
ncx = estim_params_.ncx;
@ -65,8 +63,8 @@ for i=1:npar
figunumber = figunumber+1;
hh_fig=dyn_figure(options_.nodisplay, 'Name', figurename);
end
[nam,texnam] = get_the_name(i, TeX, M_, estim_params_, options_);
[x2, f2, abscissa, dens, binf2, bsup2] = draw_prior_density(i, bayestopt_);
[nam,texnam] = get_the_name(i, TeX, M_, estim_params_, options_.varobs);
[x2, f2, ~, ~, binf2, bsup2] = draw_prior_density(i, bayestopt_);
top2 = max(f2);
if i <= nvx
name = M_.exo_names{estim_params_.var_exo(i,1)};

9
matlab/check_prior_bounds.m
View File

@ -1,10 +1,9 @@
function check_prior_bounds(xparam1,bounds,M_,estim_params_,options_,bayestopt_)
% function check_prior_bounds(xparam1,bounds,M_,estim_params_,options_)
% function check_prior_bounds(xparam1,bounds,M_,estim_params_,options_,bayestopt_)
% checks the parameter vector of violations of the prior bounds
% Inputs:
% -xparam1 [double] vector of parameters to be estimated (initial values)
% -bounds [vector] vector containing the lower and upper
% bounds
% -bounds [vector] vector containing the lower and upper bounds
% -M_ [structure] characterizing the model.
% -estim_params_ [structure] characterizing parameters to be estimated
% -options_ [structure] characterizing the options
@ -30,7 +29,7 @@ function check_prior_bounds(xparam1,bounds,M_,estim_params_,options_,bayestopt_)
outside_bound_pars=find(xparam1 < bounds.lb | xparam1 > bounds.ub);
if ~isempty(outside_bound_pars)
for ii=1:length(outside_bound_pars)
outside_bound_par_names{ii,1}=get_the_name(outside_bound_pars(ii),0,M_,estim_params_,options_);
outside_bound_par_names{ii,1}=get_the_name(outside_bound_pars(ii),0,M_,estim_params_,options_.varobs);
end
disp_string=[outside_bound_par_names{1,:}];
for ii=2:size(outside_bound_par_names,1)
@ -41,7 +40,7 @@ end
inadmissible_inverse_gamma_values=find(bayestopt_.pshape==4 & xparam1 == 0);
if ~isempty(inadmissible_inverse_gamma_values)
for ii=1:length(inadmissible_inverse_gamma_values)
inadmissible_inverse_gamma_par_names{ii,1}=get_the_name(inadmissible_inverse_gamma_values(ii),0,M_,estim_params_,options_);
inadmissible_inverse_gamma_par_names{ii,1}=get_the_name(inadmissible_inverse_gamma_values(ii),0,M_,estim_params_,options_.varobs);
end
disp_string=[inadmissible_inverse_gamma_par_names{1,:}];
for ii=2:size(inadmissible_inverse_gamma_par_names,1)

12
matlab/convergence_diagnostics/mcmc_diagnostics.m
View File

@ -71,12 +71,12 @@ param_name_tex = {};
Ifac=NaN(nblck,npar);
for jj = 1:npar
if options_.TeX
[par_name_temp, par_name_tex_temp] = get_the_name(jj, options_.TeX, M_,estim_params_, options_);
[par_name_temp, par_name_tex_temp] = get_the_name(jj, options_.TeX, M_,estim_params_, options_.varobs);
param_name = vertcat(param_name, par_name_temp);
par_name_tex_temp = strrep(par_name_tex_temp,'$','');
param_name_tex = vertcat(param_name_tex, par_name_tex_temp);
else
par_name_temp = get_the_name(jj, options_.TeX, M_, estim_params_, options_);
par_name_temp = get_the_name(jj, options_.TeX, M_, estim_params_, options_.varobs);
param_name = vertcat(param_name, par_name_temp);
end
Draws = GetAllPosteriorDraws(M_.dname, M_.fname, jj, FirstMhFile, FirstLine, TotalNumberOfMhFiles, NumberOfDraws, nblck);
@ -130,11 +130,11 @@ if options_.TeX
end
for jj=1:npar
if options_.TeX
[param_name_temp, param_name_tex_temp] = get_the_name(jj, options_.TeX, M_, estim_params_, options_);
[param_name_temp, param_name_tex_temp] = get_the_name(jj, options_.TeX, M_, estim_params_, options_.varobs);
param_name_tex = vertcat(param_name_tex, strrep(param_name_tex_temp, '$',''));
param_name = vertcat(param_name, param_name_temp);
else
param_name_temp = get_the_name(jj, options_.TeX, M_,estim_params_, options_);
param_name_temp = get_the_name(jj, options_.TeX, M_,estim_params_, options_.varobs);
param_name = vertcat(param_name, param_name_temp);
end
end
@ -290,7 +290,7 @@ for i = 1:pages
boxplot = 1;
for j = 1:npardisp % Loop over parameters %npardisp instead of 3
k = k+1;
[nam,namtex] = get_the_name(k,TeX,M_,estim_params_,options_);
[nam,namtex] = get_the_name(k,TeX,M_,estim_params_,options_.varobs);
for crit = 1:3% Loop over criteria
if crit == 1
plt1 = UDIAG(:,1,k);
@ -346,7 +346,7 @@ if reste
boxplot = 1;
for j = 1:reste
k = k+1;
[nam,namtex] = get_the_name(k,TeX,M_,estim_params_,options_);
[nam,namtex] = get_the_name(k,TeX,M_,estim_params_,options_.varobs);
for crit = 1:3
if crit == 1
plt1 = UDIAG(:,1,k);

29
matlab/dyn_latex_table.m
View File

@ -28,6 +28,25 @@ end
OutputDirectoryName = CheckPath('latex',M_.dname);
%delete dollars in label as they will be added automatically below
begin_dollar = cellfun (@(x)startsWith(x,'$'),labels,'UniformOutput',1);
end_dollar = cellfun (@(x)endsWith(x,'$'),labels,'UniformOutput',1);
if all(begin_dollar) && all(end_dollar)
labels = cellfun(@(x)delete_dollar(x,'begin'),labels,'UniformOutput',0);
labels = cellfun(@(x)delete_dollar(x,'end'),labels,'UniformOutput',0);
end
%delete dollars in headers as they will be added automatically below
begin_dollar = cellfun (@(x)startsWith(x,'$'),headers,'UniformOutput',1);
end_dollar = cellfun (@(x)endsWith(x,'$'),headers,'UniformOutput',1);
if all(begin_dollar) && all(end_dollar)
headers = cellfun(@(x)delete_dollar(x,'begin'),headers,'UniformOutput',0);
headers = cellfun(@(x)delete_dollar(x,'end'),headers,'UniformOutput',0);
end
% Set width of label column
if isempty(label_width)
label_width = cellofchararraymaxlength(vertcat(headers{1}, labels))+2;
@ -102,4 +121,12 @@ end
fprintf(fidTeX, '\\end{longtable}\n ');
fprintf(fidTeX, '\\end{center}\n');
fprintf(fidTeX, '%% End of TeX file.\n');
fclose(fidTeX);
fclose(fidTeX);
function x=delete_dollar(x,position_string)
if strcmp(position_string,'begin')
x(1)=[];
elseif strcmp(position_string,'end')
x(end)=[];
end

2
matlab/dyn_table.m
View File

@ -41,7 +41,7 @@ assert(size(data, 1) == length(rows), 'must have the same number of rows')
assert(size(data, 2) == length(cols), 'must have the same number of columns')
assert(isint(indent), 'indent must be an integer')
skipline(3)
skipline(2)
%% Print Output
rowstrlens = cellfun(@length, rows);

69
matlab/dynare_identification.m
View File

@ -1,5 +1,5 @@
function [pdraws, STO_REDUCEDFORM, STO_MOMENTS, STO_DYNAMIC, STO_si_dDYNAMIC, STO_si_dREDUCEDFORM, STO_si_dMOMENTS, STO_dSPECTRUM, STO_dMINIMAL] = dynare_identification(options_ident, pdraws0)
%function [pdraws, STO_REDUCEDFORM, STO_MOMENTS, STO_DYNAMIC, STO_si_dDYNAMIC, STO_si_dREDUCEDFORM, STO_si_dMOMENTS, STO_dSPECTRUM, STO_dMINIMAL] = dynare_identification(options_ident, pdraws0)
function [pdraws, STO_REDUCEDFORM, STO_MOMENTS, STO_DYNAMIC, STO_si_dDYNAMIC, STO_si_dREDUCEDFORM, STO_si_dMOMENTS, STO_dSPECTRUM, STO_dMINIMAL] = dynare_identification(M_,oo_,options_,bayestopt_,estim_params_,options_ident, pdraws0)
% [pdraws, STO_REDUCEDFORM, STO_MOMENTS, STO_DYNAMIC, STO_si_dDYNAMIC, STO_si_dREDUCEDFORM, STO_si_dMOMENTS, STO_dSPECTRUM, STO_dMINIMAL] = dynare_identification(options_ident, pdraws0)
% -------------------------------------------------------------------------
% This function is called, when the user specifies identification(...); in the mod file. It prepares all identification analysis:
% (1) set options, local and persistent variables for a new identification
@ -11,6 +11,11 @@ function [pdraws, STO_REDUCEDFORM, STO_MOMENTS, STO_DYNAMIC, STO_si_dDYNAMIC, ST
% to put identification in your mod file, otherwise the preprocessor won't provide all necessary objects
% =========================================================================
% INPUTS
% * M_ [structure] Matlab's structure describing the model
% * oo_ [structure] Matlab's structure describing the results
% * options_ [structure] Matlab's structure describing the current options
% * bayestopt_ [structure] describing the priors
% * estim_params_ [structure] characterizing parameters to be estimated
% * options_ident [structure] identification options
% * pdraws0 [SampleSize by totparam_nbr] optional: matrix of MC sample of model parameters
% -------------------------------------------------------------------------
@ -65,9 +70,6 @@ function [pdraws, STO_REDUCEDFORM, STO_MOMENTS, STO_DYNAMIC, STO_si_dDYNAMIC, ST
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
% =========================================================================
global M_ options_ oo_ bayestopt_ estim_params_
store_options_ = options_; % store options to restore them at the end
fname = M_.fname; %model name
dname = M_.dname; %model name
@ -255,11 +257,11 @@ if options_ident.gsa_sample_file
end
pdraws0 = [lpmatx lpmat(istable,:)];
clear lpmat lpmat0 istable;
elseif nargin==1
elseif nargin==6
pdraws0=[];
end
external_sample=0;
if nargin==2 || ~isempty(pdraws0)
if nargin==7 || ~isempty(pdraws0)
% change settings if there is an external sample provided as input argument
options_ident.prior_mc = size(pdraws0,1);
options_ident.load_ident_files = 0;
@ -361,11 +363,11 @@ if prior_exist % use estimated_params block
name_tex = cell(totparam_nbr,1); %initialize cell for TeX parameter names
for jj=1:totparam_nbr
if options_.TeX
[param_name_temp, param_name_tex_temp]= get_the_name(jj,options_.TeX,M_,estim_params_,options_);
[param_name_temp, param_name_tex_temp]= get_the_name(jj,options_.TeX,M_,estim_params_,options_.varobs);
name_tex{jj,1} =param_name_tex_temp;
name{jj,1} = param_name_temp; %ordering corresponds to estimated_params
else
param_name_temp = get_the_name(jj,options_.TeX,M_,estim_params_,options_);
param_name_temp = get_the_name(jj,options_.TeX,M_,estim_params_,options_.varobs);
name{jj,1} = param_name_temp; %ordering corresponds to estimated_params
end
end
@ -469,7 +471,7 @@ if iload <=0
options_ident.tittxt = parameters; %title text for graphs and figures
% perform identification analysis for single point
[ide_moments_point, ide_spectrum_point, ide_minimal_point, ide_hess_point, ide_reducedform_point, ide_dynamic_point, derivatives_info_point, info, error_indicator_point] = ...
identification_analysis(params, indpmodel, indpstderr, indpcorr, options_ident, dataset_info, prior_exist, 1); %the 1 at the end implies initialization of persistent variables
identification_analysis(M_,options_,oo_,bayestopt_,estim_params_,params, indpmodel, indpstderr, indpcorr, options_ident, dataset_info, prior_exist, 1); %the 1 at the end implies initialization of persistent variables
if info(1)~=0
% there are errors in the solution algorithm
message = get_error_message(info,options_);
@ -486,7 +488,7 @@ if iload <=0
options_ident.tittxt = 'Random_prior_params'; %title text for graphs and figures
% perform identification analysis
[ide_moments_point, ide_spectrum_point, ide_minimal_point, ide_hess_point, ide_reducedform_point, ide_dynamic_point, derivatives_info_point, info, error_indicator_point] = ...
identification_analysis(params, indpmodel, indpstderr, indpcorr, options_ident, dataset_info, prior_exist, 1);
identification_analysis(M_,options_,oo_,bayestopt_,estim_params_,params, indpmodel, indpstderr, indpcorr, options_ident, dataset_info, prior_exist, 1);
end
end
if info(1)
@ -514,7 +516,8 @@ if iload <=0
disp_identification(params, ide_reducedform_point, ide_moments_point, ide_spectrum_point, ide_minimal_point, name, options_ident);
if ~options_ident.no_identification_strength && ~options_.nograph && ~error_indicator_point.identification_strength && ~error_indicator_point.identification_moments
% plot (i) identification strength and sensitivity measure based on the moment information matrix and (ii) plot advanced analysis graphs
plot_identification(params, ide_moments_point, ide_hess_point, ide_reducedform_point, ide_dynamic_point, options_ident.advanced, parameters, name, IdentifDirectoryName, parameters_TeX, name_tex);
plot_identification(M_,params, ide_moments_point, ide_hess_point, ide_reducedform_point, ide_dynamic_point, options_ident.advanced, parameters, name, ...
IdentifDirectoryName, M_.fname, options_, estim_params_, bayestopt_, parameters_TeX, name_tex);
end
if SampleSize > 1
@ -540,7 +543,7 @@ if iload <=0
options_ident.tittxt = []; % clear title text for graphs and figures
% run identification analysis
[ide_moments, ide_spectrum, ide_minimal, ide_hess, ide_reducedform, ide_dynamic, ide_derivatives_info, info, error_indicator] = ...
identification_analysis(params, indpmodel, indpstderr, indpcorr, options_MC, dataset_info, prior_exist, 0); % the 0 implies that we do not initialize persistent variables anymore
identification_analysis(M_,options_,oo_,bayestopt_,estim_params_,params, indpmodel, indpstderr, indpcorr, options_MC, dataset_info, prior_exist, 0); % the 0 implies that we do not initialize persistent variables anymore
if iteration==0 && info(1)==0 % preallocate storage in the first admissable run
delete([IdentifDirectoryName '/' fname '_identif_*.mat']) % delete previously saved results
@ -783,6 +786,19 @@ if iload <=0
else
maxrun_dMINIMAL = 0;
end
si_dDYNAMICnorm=NaN(max([maxrun_dDYNAMIC, maxrun_dREDUCEDFORM, maxrun_dMOMENTS, maxrun_dSPECTRUM, maxrun_dMINIMAL]),size(STO_si_dDYNAMIC,2));
if ~options_MC.no_identification_reducedform
si_dREDUCEDFORMnorm=NaN(max([maxrun_dDYNAMIC, maxrun_dREDUCEDFORM, maxrun_dMOMENTS, maxrun_dSPECTRUM, maxrun_dMINIMAL]),size(STO_si_dREDUCEDFORM,2));
end
if ~options_MC.no_identification_moments
si_dMOMENTSnorm=NaN(max([maxrun_dDYNAMIC, maxrun_dREDUCEDFORM, maxrun_dMOMENTS, maxrun_dSPECTRUM, maxrun_dMINIMAL]),size(STO_si_dMOMENTS,2));
end
if ~options_MC.no_identification_spectrum
dSPECTRUMnorm=NaN(max([maxrun_dDYNAMIC, maxrun_dREDUCEDFORM, maxrun_dMOMENTS, maxrun_dSPECTRUM, maxrun_dMINIMAL]),size(STO_dSPECTRUM,2));
end
if ~options_MC.no_identification_minimal
dMINIMALnorm=NaN(max([maxrun_dDYNAMIC, maxrun_dREDUCEDFORM, maxrun_dMOMENTS, maxrun_dSPECTRUM, maxrun_dMINIMAL]),size(STO_dMINIMAL,2));
end
for irun=1:max([maxrun_dDYNAMIC, maxrun_dREDUCEDFORM, maxrun_dMOMENTS, maxrun_dSPECTRUM, maxrun_dMINIMAL])
iter=iter+1;
% note that this is not the same si_dDYNAMICnorm as computed in identification_analysis
@ -863,7 +879,8 @@ if iload
disp_identification(ide_hess_point.params, ide_reducedform_point, ide_moments_point, ide_spectrum_point, ide_minimal_point, name, options_ident);
if ~options_.nograph && ~error_indicator_point.identification_strength && ~error_indicator_point.identification_moments
% plot (i) identification strength and sensitivity measure based on the sample information matrix and (ii) advanced analysis graphs
plot_identification(ide_hess_point.params, ide_moments_point, ide_hess_point, ide_reducedform_point, ide_dynamic_point, options_ident.advanced, parameters, name, IdentifDirectoryName, [], name_tex);
plot_identification(M_,ide_hess_point.params, ide_moments_point, ide_hess_point, ide_reducedform_point, ide_dynamic_point, options_ident.advanced, parameters, name, ...
IdentifDirectoryName, M_.fname, options_, estim_params_, bayestopt_, [], name_tex);
end
end
@ -877,7 +894,8 @@ if SampleSize > 1
options_ident.advanced = advanced0; % reset advanced setting
if ~options_.nograph && isfield(ide_hess_point,'ide_strength_dMOMENTS')
% plot (i) identification strength and sensitivity measure based on the sample information matrix and (ii) advanced analysis graphs
plot_identification(pdraws, IDE_MOMENTS, ide_hess_point, IDE_REDUCEDFORM, IDE_DYNAMIC, options_ident.advanced, 'MC sample ', name, IdentifDirectoryName, [], name_tex);
plot_identification(M_, pdraws, IDE_MOMENTS, ide_hess_point, IDE_REDUCEDFORM, IDE_DYNAMIC, options_ident.advanced, 'MC sample ', name, ...
IdentifDirectoryName, M_.fname, options_, estim_params_, bayestopt_, [], name_tex);
end
%advanced display and plots for MC Sample, i.e. look at draws with highest/lowest condition number
if options_ident.advanced
@ -894,7 +912,7 @@ if SampleSize > 1
if ~iload
options_ident.tittxt = tittxt; %title text for graphs and figures
[ide_moments_max, ide_spectrum_max, ide_minimal_max, ide_hess_max, ide_reducedform_max, ide_dynamic_max, derivatives_info_max, info_max, error_indicator_max] = ...
identification_analysis(pdraws(jmax,:), indpmodel, indpstderr, indpcorr, options_ident, dataset_info, prior_exist, 1); %the 1 at the end initializes some persistent variables
identification_analysis(M_,options_,oo_,bayestopt_,estim_params_,pdraws(jmax,:), indpmodel, indpstderr, indpcorr, options_ident, dataset_info, prior_exist, 1); %the 1 at the end initializes some persistent variables
save([IdentifDirectoryName '/' fname '_identif.mat'], 'ide_hess_max', 'ide_moments_max', 'ide_spectrum_max', 'ide_minimal_max','ide_reducedform_max', 'ide_dynamic_max', 'jmax', '-append');
end
advanced0 = options_ident.advanced; options_ident.advanced = 1; % make sure advanced setting is on
@ -902,7 +920,8 @@ if SampleSize > 1
options_ident.advanced = advanced0; %reset advanced setting
if ~options_.nograph && ~error_indicator_max.identification_strength && ~error_indicator_max.identification_moments
% plot (i) identification strength and sensitivity measure based on the sample information matrix and (ii) advanced analysis graphs
plot_identification(pdraws(jmax,:), ide_moments_max, ide_hess_max, ide_reducedform_max, ide_dynamic_max, 1, tittxt, name, IdentifDirectoryName, tittxt, name_tex);
plot_identification(M_, pdraws(jmax,:), ide_moments_max, ide_hess_max, ide_reducedform_max, ide_dynamic_max, 1, tittxt, name, ...
IdentifDirectoryName, M_.fname, options_, estim_params_, bayestopt_, tittxt, name_tex);
end
% SMALLEST condition number
@ -911,8 +930,8 @@ if SampleSize > 1
fprintf('Testing %s.\n',tittxt);
if ~iload
options_ident.tittxt = tittxt; %title text for graphs and figures
[ide_moments_min, ide_spectrum_min, ide_minimal_min, ide_hess_min, ide_reducedform_min, ide_dynamic_min, derivatives_info_min, info_min, error_indicator_min] = ...
identification_analysis(pdraws(jmin,:), indpmodel, indpstderr, indpcorr, options_ident, dataset_info, prior_exist, 1); %the 1 at the end initializes persistent variables
[ide_moments_min, ide_spectrum_min, ide_minimal_min, ide_hess_min, ide_reducedform_min, ide_dynamic_min, ~, ~, error_indicator_min] = ...
identification_analysis(M_,options_,oo_,bayestopt_,estim_params_,pdraws(jmin,:), indpmodel, indpstderr, indpcorr, options_ident, dataset_info, prior_exist, 1); %the 1 at the end initializes persistent variables
save([IdentifDirectoryName '/' fname '_identif.mat'], 'ide_hess_min', 'ide_moments_min','ide_spectrum_min','ide_minimal_min','ide_reducedform_min', 'ide_dynamic_min', 'jmin', '-append');
end
advanced0 = options_ident.advanced; options_ident.advanced = 1; % make sure advanced setting is on
@ -920,7 +939,8 @@ if SampleSize > 1
options_ident.advanced = advanced0; %reset advanced setting
if ~options_.nograph && ~error_indicator_min.identification_strength && ~error_indicator_min.identification_moments
% plot (i) identification strength and sensitivity measure based on the sample information matrix and (ii) advanced analysis graphs
plot_identification(pdraws(jmin,:),ide_moments_min,ide_hess_min,ide_reducedform_min,ide_dynamic_min,1,tittxt,name,IdentifDirectoryName,tittxt,name_tex);
plot_identification(M_, pdraws(jmin,:),ide_moments_min,ide_hess_min,ide_reducedform_min,ide_dynamic_min,1,tittxt,name,...
IdentifDirectoryName, M_.fname, options_, estim_params_, bayestopt_, tittxt,name_tex);
end
% reset nodisplay option
options_.nodisplay = store_nodisplay;
@ -934,14 +954,15 @@ if SampleSize > 1
if ~iload
options_ident.tittxt = tittxt; %title text for graphs and figures
[ide_moments_(j), ide_spectrum_(j), ide_minimal_(j), ide_hess_(j), ide_reducedform_(j), ide_dynamic_(j), derivatives_info_(j), info_resolve, error_indicator_j] = ...
identification_analysis(pdraws(jcrit(j),:), indpmodel, indpstderr, indpcorr, options_ident, dataset_info, prior_exist, 1);
identification_analysis(M_,options_,oo_,bayestopt_,estim_params_,pdraws(jcrit(j),:), indpmodel, indpstderr, indpcorr, options_ident, dataset_info, prior_exist, 1);
end
advanced0 = options_ident.advanced; options_ident.advanced = 1; %make sure advanced setting is on
disp_identification(pdraws(jcrit(j),:), ide_reducedform_(j), ide_moments_(j), ide_spectrum_(j), ide_minimal_(j), name, options_ident);
options_ident.advanced = advanced0; % reset advanced
if ~options_.nograph && ~error_indicator_j.identification_strength && ~error_indicator_j.identification_moments
% plot (i) identification strength and sensitivity measure based on the sample information matrix and (ii) advanced analysis graphs
plot_identification(pdraws(jcrit(j),:), ide_moments_(j), ide_hess_(j), ide_reducedform_(j), ide_dynamic_(j), 1, tittxt, name, IdentifDirectoryName, tittxt, name_tex);
plot_identification(M_, pdraws(jcrit(j),:), ide_moments_(j), ide_hess_(j), ide_reducedform_(j), ide_dynamic_(j), 1, tittxt, name, ...
IdentifDirectoryName, M_.fname, options_, estim_params_, bayestopt_, tittxt, name_tex);
end
end
if ~iload
@ -957,6 +978,4 @@ end
%reset warning state
warning_config;
fprintf('\n==== Identification analysis completed ====\n\n')
options_ = store_options_; %restore options set
fprintf('\n==== Identification analysis completed ====\n\n')

164
matlab/dynare_sensitivity.m
View File

@ -1,10 +1,19 @@
function x0=dynare_sensitivity(options_gsa)
function x0=dynare_sensitivity(M_,oo_,options_,bayestopt_,estim_params_,options_gsa)
% x0=dynare_sensitivity(M_,oo_,options_,bayestopt_,estim_params_,options_gsa)
% Frontend to the Sensitivity Analysis Toolbox for DYNARE
% Inputs:
% - M_ [structure] Matlab's structure describing the model
% - oo_ [structure] Matlab's structure describing the results
% - options_ [structure] Matlab's structure describing the current options
% - bayestopt_ [structure] describing the priors
% - estim_params_ [structure] characterizing parameters to be estimated
% - options_gsa [structure] Matlab's structure describing the GSA options
%
% Reference:
% M. Ratto, Global Sensitivity Analysis for Macroeconomic models, MIMEO, 2006.
% M. Ratto (2008), Analysing DSGE Models with Global Sensitivity Analysis,
% Computational Economics (2008), 31, pp. 115–139
% Copyright © 2008-2018 Dynare Team
% Copyright © 2008-2023 Dynare Team
%
% This file is part of Dynare.
%
@ -21,14 +30,11 @@ function x0=dynare_sensitivity(options_gsa)
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
global M_ options_ oo_ bayestopt_ estim_params_
if options_.dsge_var
error('Identification does not support DSGE-VARs at the current stage')
end
fname_ = M_.fname;
lgy_ = M_.endo_names;
x0=[];
% check user defined options
@ -43,7 +49,6 @@ end
if isfield(options_gsa,'morris') && options_gsa.morris==1
if isfield(options_gsa,'identification') && options_gsa.identification==0
% options_gsa.redform=1;
end
if isfield(options_gsa,'ppost') && options_gsa.ppost
error('sensitivity:: Morris is incompatible with posterior sampling')
@ -89,9 +94,6 @@ if options_.order~=1
options_.order = 1;
end
original_prior_trunc = options_.prior_trunc;
original_qz_criterium = options_.qz_criterium;
if ~isempty(options_gsa.datafile) || isempty(bayestopt_) || options_gsa.rmse
if isempty(options_gsa.datafile) && options_gsa.rmse
disp('The data file and all relevant estimation options ')
@ -130,7 +132,7 @@ if ~isempty(options_gsa.datafile) || isempty(bayestopt_) || options_gsa.rmse
options_.mode_compute = 0;
options_.filtered_vars = 1;
options_.plot_priors = 0;
[dataset_,dataset_info,xparam1,hh, M_, options_, oo_, estim_params_, bayestopt_] = ...
[dataset_,dataset_info,~,~, M_, options_, oo_, estim_params_, bayestopt_] = ...
dynare_estimation_init(M_.endo_names, fname_, 1, M_, options_, oo_, estim_params_, bayestopt_);
% computes a first linear solution to set up various variables
else
@ -146,7 +148,7 @@ if M_.exo_nbr==0
error('dynare_sensitivity does not support having no varexo in the model. As a workaround you could define a dummy exogenous variable.')
end
[make,my,day,punk,oo_.dr,M_.params] = dynare_resolve(M_,options_,oo_.dr,oo_.steady_state,oo_.exo_steady_state,oo_.exo_det_steady_state);
[~,~,~,~,oo_.dr,M_.params] = dynare_resolve(M_,options_,oo_.dr,oo_.steady_state,oo_.exo_steady_state,oo_.exo_det_steady_state);
options_gsa = set_default_option(options_gsa,'identification',0);
if options_gsa.identification
@ -166,7 +168,6 @@ if options_gsa.identification
options_.options_ident.load_ident_files = options_gsa.load_ident_files;
options_.options_ident.useautocorr = options_gsa.useautocorr;
options_.options_ident.ar = options_gsa.ar;
options_ident=options_.options_ident;
else
options_ident=[];
options_ident = set_default_option(options_ident,'load_ident_files',options_gsa.load_ident_files);
@ -193,7 +194,6 @@ options_gsa = set_default_option(options_gsa,'load_stab',0);
options_gsa = set_default_option(options_gsa,'alpha2_stab',0);
options_gsa = set_default_option(options_gsa,'pvalue_ks',0.001);
options_gsa = set_default_option(options_gsa,'pvalue_corr',1.e-5);
%options_gsa = set_default_option(options_gsa,'load_mh',0);
% REDFORM mapping
options_gsa = set_default_option(options_gsa,'redform',0);
options_gsa = set_default_option(options_gsa,'load_redform',0);
@ -202,8 +202,29 @@ options_gsa = set_default_option(options_gsa,'threshold_redform',[]);
options_gsa = set_default_option(options_gsa,'ksstat_redform',0.001);
options_gsa = set_default_option(options_gsa,'alpha2_redform',1.e-5);
options_gsa = set_default_option(options_gsa,'namendo',{});
options_gsa = set_default_option(options_gsa,'namlagendo',[]);
options_gsa = set_default_option(options_gsa,'namlagendo',{});
options_gsa = set_default_option(options_gsa,'namexo',{});
options_gsa = set_default_option(options_gsa,'namendo_tex',{});
options_gsa = set_default_option(options_gsa,'namlagendo_tex',{});
options_gsa = set_default_option(options_gsa,'namexo_tex',{});
if strmatch(':',options_gsa.namendo,'exact')
options_gsa.namendo = M_.endo_names(1:M_.orig_endo_nbr);
end
if strmatch(':',options_gsa.namexo,'exact')
options_gsa.namexo = M_.exo_names;
end
if strmatch(':',options_gsa.namlagendo,'exact')
options_gsa.namlagendo = M_.endo_names(1:M_.orig_endo_nbr);
end
if options_.TeX
[~,Locb]=ismember(options_gsa.namendo,M_.endo_names);
options_gsa.namendo_tex=cellfun(@(x) horzcat('$', x, '$'), M_.endo_names_tex(Locb), 'UniformOutput', false);
[~,Locb]=ismember(options_gsa.namlagendo,M_.endo_names);
options_gsa.namlagendo_tex=cellfun(@(x) horzcat('$', x, '$'), M_.endo_names_tex(Locb), 'UniformOutput', false);
[~,Locb]=ismember(options_gsa.namexo,M_.exo_names);
options_gsa.namexo_tex=cellfun(@(x) horzcat('$', x, '$'), M_.exo_names_tex(Locb), 'UniformOutput', false);
end
% RMSE mapping
options_gsa = set_default_option(options_gsa,'load_rmse',0);
options_gsa = set_default_option(options_gsa,'lik_only',0);
@ -212,8 +233,9 @@ options_gsa = set_default_option(options_gsa,'var_rmse', options_.varobs);
options_gsa.var_rmse_tex={};
for ii=1:length(options_gsa.var_rmse)
temp_name = M_.endo_names_tex{strmatch(options_gsa.var_rmse{ii}, M_.endo_names, 'exact')};
options_gsa.var_rmse_tex = vertcat(options_gsa.var_rmse_tex, temp_name);
options_gsa.var_rmse_tex = vertcat(options_gsa.var_rmse_tex, ['$' temp_name '$']);
end
options_gsa.varobs_tex = cellfun(@(x) horzcat('$', x, '$'), M_.endo_names_tex(options_.varobs_id), 'UniformOutput', false);
options_gsa = set_default_option(options_gsa,'pfilt_rmse', 0.1);
options_gsa = set_default_option(options_gsa,'istart_rmse', options_.presample+1);
options_gsa = set_default_option(options_gsa,'alpha_rmse', 0.001);
@ -245,31 +267,22 @@ if options_gsa.morris==1
options_gsa.pprior=1;
end
options_gsa.ppost=0;
%options_gsa.stab=1;
options_gsa.glue=0;
options_gsa.rmse=0;
options_gsa.load_rmse=0;
options_gsa.alpha2_stab=1;
options_gsa.pvalue_ks=0;
options_gsa.pvalue_corr=0;
% if options_gsa.morris==3,
% options_gsa = set_default_option(options_gsa,'Nsam',256);
% OutputDirectoryName = CheckPath('gsa/identif',M_.dname);
% else
OutputDirectoryName = CheckPath('gsa/screen',M_.dname);
% end
else
OutputDirectoryName = CheckPath('gsa',M_.dname);
end
% options_.opt_gsa = options_gsa;
if (options_gsa.load_stab || options_gsa.load_rmse || options_gsa.load_redform) && options_gsa.pprior
filetoload=[OutputDirectoryName '/' fname_ '_prior.mat'];
if ~exist(filetoload,'file')
disp([filetoload,' not found!'])
disp(['You asked to load a non existent analysis'])
%options_gsa.load_stab=0;
disp('You asked to load a non existent analysis')
return
else
if isempty(strmatch('bkpprior',who('-file', filetoload),'exact'))
@ -293,7 +306,7 @@ if (options_gsa.load_stab || options_gsa.load_rmse || options_gsa.load_redform)
end
if options_gsa.stab && ~options_gsa.ppost
x0 = stab_map_(OutputDirectoryName,options_gsa);
x0 = stab_map_(OutputDirectoryName,options_gsa,M_,oo_,options_,bayestopt_,estim_params_);
if isempty(x0)
skipline()
disp('Sensitivity computations stopped: no parameter set provided a unique solution')
@ -301,16 +314,13 @@ if options_gsa.stab && ~options_gsa.ppost
end
end
% reduced form
% redform_map(namendo, namlagendo, namexo, icomp, pprior, ilog, threshold)
options_.opt_gsa = options_gsa;
if ~isempty(options_gsa.moment_calibration) || ~isempty(options_gsa.irf_calibration)
map_calibration(OutputDirectoryName, M_, options_, oo_, estim_params_,bayestopt_);
end
if options_gsa.identification
map_ident_(OutputDirectoryName,options_gsa);
map_ident_(OutputDirectoryName,options_gsa,M_,oo_,options_,estim_params_,bayestopt_);
end
if options_gsa.redform && ~isempty(options_gsa.namendo)
@ -318,7 +328,7 @@ if options_gsa.redform && ~isempty(options_gsa.namendo)
filnam = dir([M_.dname filesep 'metropolis' filesep '*param_irf*.mat']);
lpmat=[];
for j=1:length(filnam)
load ([M_.dname filesep 'metropolis' filesep M_.fname '_param_irf' int2str(j) '.mat'])
load ([M_.dname filesep 'metropolis' filesep M_.fname '_param_irf' int2str(j) '.mat'],'stock')
lpmat=[lpmat; stock];
end
clear stock
@ -336,20 +346,10 @@ if options_gsa.redform && ~isempty(options_gsa.namendo)
save([OutputDirectoryName filesep M_.fname '_mc.mat'],'lpmat','lpmat0','istable','iunstable','iwrong','iindeterm')
options_gsa.load_stab=1;
x0 = stab_map_(OutputDirectoryName,options_gsa);
x0 = stab_map_(OutputDirectoryName,options_gsa,M_,oo_,options_,bayestopt_,estim_params_);
end
if strmatch(':',options_gsa.namendo,'exact')
options_gsa.namendo = M_.endo_names(1:M_.orig_endo_nbr);
end
if strmatch(':',options_gsa.namexo,'exact')
options_gsa.namexo = M_.exo_names;
end
if strmatch(':',options_gsa.namlagendo,'exact')
options_gsa.namlagendo = M_.endo_names(1:M_.orig_endo_nbr);
end
% options_.opt_gsa = options_gsa;
if options_gsa.morris==1
redform_screen(OutputDirectoryName,options_gsa);
redform_screen(OutputDirectoryName,options_gsa, estim_params_, M_, oo_.dr, options_, bayestopt_);
else
% check existence of the SS_ANOVA toolbox
if isempty(options_gsa.threshold_redform) && ~(exist('gsa_sdp','file')==6 || exist('gsa_sdp','file')==2)
@ -360,11 +360,10 @@ if options_gsa.redform && ~isempty(options_gsa.namendo)
fprintf('After obtaining the files, you need to unpack them and set a Matlab Path to those files.\n')
error('SS-ANOVA-R Toolbox missing!')
end
redform_map(OutputDirectoryName,options_gsa);
redform_map(OutputDirectoryName,options_gsa,M_,estim_params_,options_,bayestopt_,oo_);
end
end
% RMSE mapping
% function [rmse_MC, ixx] = filt_mc_(vvarvecm, loadSA, pfilt, alpha, alpha2)
options_.opt_gsa = options_gsa;
if options_gsa.rmse
if ~options_gsa.ppost
@ -391,7 +390,6 @@ if options_gsa.rmse
options_.forecast=0;
options_.filtered_vars=0;
end
% dynare_MC([],OutputDirectoryName,data,rawdata,data_info);
if options_gsa.pprior
TmpDirectoryName = ([M_.dname filesep 'gsa' filesep 'prior']);
else
@ -408,37 +406,18 @@ if options_gsa.rmse
delete([TmpDirectoryName filesep filparam(j).name]);
end
end
end
oo_=prior_posterior_statistics('gsa',dataset_, dataset_info,M_,oo_,options_,estim_params_,bayestopt_,'gsa::mcmc');
if options_.bayesian_irf
oo_=PosteriorIRF('gsa',options_,estim_params_,oo_,M_,bayestopt_,dataset_,dataset_info,'gsa::mcmc');
end
options_gsa.load_rmse=0;
% else
% if options_gsa.load_rmse==0,
% disp('You already saved a MC filter/smoother analysis ')
% disp('Do you want to overwrite ?')
% pause;
% if options_gsa.pprior
% delete([OutputDirectoryName,'/',fname_,'_prior_*.mat'])
% else
% delete([OutputDirectoryName,'/',fname_,'_mc_*.mat'])
% end
% dynare_MC([],OutputDirectoryName);
% options_gsa.load_rmse=0;
% end
end
end
clear a;
% filt_mc_(OutputDirectoryName,data_info);
filt_mc_(OutputDirectoryName,options_gsa,dataset_,dataset_info);
filt_mc_(OutputDirectoryName,options_gsa,dataset_,dataset_info,M_,oo_,options_,bayestopt_,estim_params_);
end
options_.opt_gsa = options_gsa;
options_.prior_trunc=original_prior_trunc;
options_.qz_criterium=original_qz_criterium ;
if options_gsa.glue
dr_ = oo_.dr;
@ -453,11 +432,10 @@ if options_gsa.glue
end
end
if ~exist('x','var')
disp(['No RMSE analysis is available for current options'])
disp(['No GLUE file prepared'])
disp('No RMSE analysis is available for current options')
disp('No GLUE file prepared')
return,
end
nruns=size(x,1);
gend = options_.nobs;
rawdata = read_variables(options_.datafile,options_.varobs,[],options_.xls_sheet,options_.xls_range);
rawdata = rawdata(options_.first_obs:options_.first_obs+gend-1,:);
@ -465,28 +443,20 @@ if options_gsa.glue
rawdata = log(rawdata);
end
if options_.prefilter == 1
%data = transpose(rawdata-ones(gend,1)*bayestopt_.mean_varobs);
data = transpose(rawdata-ones(gend,1)*mean(rawdata,1));
else
data = transpose(rawdata);
end
Obs.data = data;
Obs.time = [1:gend];
Obs.time = 1:gend;
Obs.num = gend;
for j=1:length(options_.varobs)
Obs.name{j} = options_.varobs{j};
vj = options_.varobs{j};
jxj = strmatch(vj,lgy_(dr_.order_var),'exact');
js = strmatch(vj,lgy_,'exact');
jxj = strmatch(vj,M_.endo_names(dr_.order_var),'exact');
if ~options_gsa.ppost
% y0=zeros(gend+1,nruns);
% nb = size(stock_filter,3);
% y0 = squeeze(stock_filter(:,jxj,:)) + ...
% kron(stock_ys(js,:),ones(size(stock_filter,1),1));
% Out(j).data = y0';
% Out(j).time = [1:size(y0,1)];
Out(j).data = jxj;
Out(j).time = [pwd,'/',OutputDirectoryName];
else
@ -500,17 +470,7 @@ if options_gsa.glue
Lik(j).isam = 1;
Lik(j).data = rmse_MC(:,j)';
if ~options_gsa.ppost
% y0 = squeeze(stock_smooth(:,jxj,:)) + ...
% kron(stock_ys(js,:),ones(size(stock_smooth,1),1));
% Out1(j).name = vj;
% Out1(j).ini = 'yes';
% Out1(j).time = [1:size(y0,1)];
% Out1(j).data = y0';
Out1=Out;
else
Out1=Out;
end
Out1=Out;
ismoo(j)=jxj;
end
@ -520,10 +480,6 @@ if options_gsa.glue
jsmoo=jsmoo+1;
vj = M_.endo_names{dr_.order_var(j)};
if ~options_gsa.ppost
% y0 = squeeze(stock_smooth(:,j,:)) + ...
% kron(stock_ys(j,:),ones(size(stock_smooth,1),1));
% Out1(jsmoo).time = [1:size(y0,1)];
% Out1(jsmoo).data = y0';
Out1(jsmoo).data = j;
Out1(jsmoo).time = [pwd,'/',OutputDirectoryName];
else
@ -546,36 +502,24 @@ if options_gsa.glue
end
Sam.name = bayestopt_.name;
Sam.dim = [size(x) 0];
Sam.data = [x];
Sam.data = x;
Rem.id = 'Original';
Rem.ind= [1:size(x,1)];
Rem.ind= 1:size(x,1);
Info.dynare=M_.fname;
Info.order_var=dr_.order_var;
Out=Out1;
if options_gsa.ppost
% Info.dynare=M_.fname;
% Info.order_var=dr_.order_var;
% Out=Out1;
Info.TypeofSample='post';
save([OutputDirectoryName,'/',fname_,'_glue_post.mat'], 'Out', 'Sam', 'Lik', 'Obs', 'Rem','Info', 'Exo')
%save([fname_,'_post_glue_smooth'], 'Out', 'Sam', 'Lik', 'Obs', 'Rem','Info')
else
if options_gsa.pprior
Info.TypeofSample='prior';
save([OutputDirectoryName,'/',fname_,'_glue_prior.mat'], 'Out', 'Sam', 'Lik', 'Obs', 'Rem','Info', 'Exo')
% save([OutputDirectoryName,'/',fname_,'_prior_glue'], 'Out', 'Sam', 'Lik', 'Obs', 'Rem')
% Out=Out1;
% save([OutputDirectoryName,'/',fname_,'_prior_glue_smooth'], 'Out', 'Sam', 'Lik', 'Obs', 'Rem')
else
Info.TypeofSample='mc';
save([OutputDirectoryName,'/',fname_,'_glue_mc.mat'], 'Out', 'Sam', 'Lik', 'Obs', 'Rem','Info', 'Exo')
% save([OutputDirectoryName,'/',fname_,'_mc_glue'], 'Out', 'Sam', 'Lik', 'Obs', 'Rem')
% Out=Out1;
% save([OutputDirectoryName,'/',fname_,'_mc_glue_smooth'], 'Out', 'Sam', 'Lik', 'Obs', 'Rem')
end
end
end
end

4
matlab/fjaco.m
View File

@ -1,5 +1,5 @@
function fjac = fjaco(f,x,varargin)
% fjac = fjaco(f,x,varargin)
% FJACO Computes two-sided finite difference Jacobian
% USAGE
% fjac = fjaco(f,x,P1,P2,...)
@ -57,7 +57,7 @@ feval(f,x,varargin{:});
%Auxiliary functions
function disp_info_error_identification_perturbation(info,j)
% there are errors in the solution algorithm
probl_par = get_the_name(j,varargin{4}.TeX,varargin{3},varargin{2},varargin{4});
probl_par = get_the_name(j,varargin{4}.TeX,varargin{3},varargin{2},varargin{4}.varobs);
skipline()
message = get_error_message(info,varargin{4});
fprintf('Parameter error in numerical two-sided difference method:\n')

67
matlab/get_the_name.m
View File

@ -1,46 +1,26 @@
function [nam, texnam] = get_the_name(k, TeX, M_, estim_params_, options_)
function [nam, texnam] = get_the_name(k, TeX, M_, estim_params_, varobs)
% [nam, texnam] = get_the_name(k, TeX, M_, estim_params_, varobs)
% Returns name of estimated parameter number k, following the internal ordering of
% the estimated parameters.
% Inputs:
% - k [integer] parameter number.
% - TeX [bool] if false, texnam is not returned (empty matrix)
% - M_ [structure] model
% - estim_params_ [structure] describing the estimated parameters
% - varobs [cell] name of observed variables
%
% Outputs
% - nam [char] internal name of the variable
% - texnam [char] TeX name of the same variable (if defined in the mod file)
%
% This function is called by:
% get_prior_info, mcmc_diagnostics, mode_check, PlotPosteriorDistributions, plot_priors
%
% This function calls:
% None.
%
%@info:
%! @deftypefn {Function File} {[@var{nam},@var{texnam}] =} get_the_name (@var{k},@var{TeX},@var{M_},@var{estim_params_},@var{options_})
%! @anchor{get_the_name}
%! @sp 1
%! Returns the name of the estimated parameter number @var{k}, following the internal ordering of the estimated parameters.
%! @sp 2
%! @strong{Inputs}
%! @sp 1
%! @table @ @var
%! @item k
%! Integer scalar, parameter number.
%! @item TeX
%! Integer scalar, if @var{TeX}==0 then @var{texnam} is not returned (empty matrix).
%! @item M_
%! Matlab's structure describing the model (initialized by @code{dynare}).
%! @item estim_params_
%! Matlab's structure describing the estimated parameters (initialized by @code{dynare}).
%! @item options_
%! Matlab's structure describing the options (initialized by @code{dynare}).
%! @end table
%! @sp 2
%! @strong{Outputs}
%! @sp 1
%! @table @ @var
%! @item nam
%! String, internal name of the variable
%! @item texnam
%! String, TeX name of the same variable (if defined in the mod file).
%! @end table
%! @sp 2
%! @strong{This function is called by:}
%! @sp 1
%! @ref{get_prior_info}, @ref{mcmc_diagnostics}, @ref{mode_check}, @ref{PlotPosteriorDistributions}, @ref{plot_priors}
%! @sp 2
%! @strong{This function calls:}
%! @sp 1
%! None.
%! @end deftypefn
%@eod:
% Copyright © 2004-2018 Dynare Team
% Copyright © 2004-2023 Dynare Team
%
% This file is part of Dynare.
%
@ -57,7 +37,6 @@ function [nam, texnam] = get_the_name(k, TeX, M_, estim_params_, options_)
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
nam = [];
texnam = [];
nvx = estim_params_.nvx;
@ -73,7 +52,7 @@ if k <= nvx
texnam = sprintf('$ \\sigma_{%s} $', tname);
end
elseif k <= (nvx+nvn)
vname = options_.varobs{estim_params_.nvn_observable_correspondence(k-estim_params_.nvx,1)};
vname = varobs{estim_params_.nvn_observable_correspondence(k-estim_params_.nvx,1)};
nam = sprintf('SE_EOBS_%s', vname);
if TeX
tname = M_.endo_names_tex{estim_params_.var_endo(k-estim_params_.nvx,1)};

46
matlab/gsa/Morris_Measure_Groups.m
View File

@ -6,20 +6,20 @@ function [SAmeas, OutMatrix] = Morris_Measure_Groups(NumFact, Sample, Output, p,
% INPUTS
% -------------------------------------------------------------------------
% Group [NumFactor, NumGroups] := Matrix describing the groups.
% Each column represents one group.
% The element of each column are zero if the factor is not in the
% group. Otherwise it is 1.
% Each column represents one group.
% The element of each column are zero if the factor is not in the
% group. Otherwise it is 1.
%
% Sample := Matrix of the Morris sampled trajectories
%
% Output := Matrix of the output(s) values in correspondence of each point
% of each trajectory
% of each trajectory
%
% k = Number of factors
% -------------------------------------------------------------------------
% OUTPUTS
% OutMatrix (NumFactor*NumOutputs, 3)= [Mu*, Mu, StDev]
% for each output it gives the three measures of each factor
% for each output it gives the three measures of each factor
% -------------------------------------------------------------------------
%
% Written by Jessica Cariboni and Francesca Campolongo
@ -27,7 +27,7 @@ function [SAmeas, OutMatrix] = Morris_Measure_Groups(NumFact, Sample, Output, p,
%
% Copyright © 2005 European Commission
% Copyright © 2012-2017 Dynare Team
% Copyright © 2012-2023 Dynare Team
%
% This file is part of Dynare.
%
@ -50,32 +50,34 @@ if nargin==0
return
end
OutMatrix=[];
if nargin < 5, Group=[]; end
NumGroups = size(Group,2);
if nargin < 4 | isempty(p)
if nargin < 4 || isempty(p)
p = 4;
end
Delt = p/(2*p-2);
if NumGroups ~ 0
if NumGroups ~= 0
sizea = NumGroups; % Number of groups
GroupMat=Group;
GroupMat = GroupMat';
else
sizea = NumFact;
end
r=size(Sample,1)/(sizea+1); % Number of trajectories
if NumGroups > 0
OutMatrix = NaN(sizea*size(Output,2),1);
else
OutMatrix = NaN(sizea*size(Output,2),3);
end
% For Each Output
for k=1:size(Output,2)
OutValues=Output(:,k);
% For each r trajectory
for i=1:r
% For each step j in the trajectory
% Read the orientation matrix fact for the r-th sampling
% Read the corresponding output values
@ -86,13 +88,11 @@ for k=1:size(Output,2)
% For each point of the fixed trajectory compute the values of the Morris function. The function
% is partitioned in four parts, from order zero to order 4th.
% SAmeas=NaN();
for j=1:sizea % For each point in the trajectory i.e for each factor
% matrix of factor which changes
if NumGroups ~ 0
if NumGroups ~= 0
AuxFind (:,1) = A(:,j);
% AuxFind(find(A(:,j)),1)=1;
% Pippo = sum((Group - repmat(AuxFind,1,NumGroups)),1);
% Change_factor(j,i) = find(Pippo==0);
Change_factor = find(abs(AuxFind)>1e-010);
% If we deal with groups we can only estimate the new mu*
% measure since factors in the same groups can move in
@ -100,7 +100,6 @@ for k=1:size(Output,2)
% Morris mu cannopt be applied.
% In the new version the elementary effect is defined with
% the absolute value.
%SAmeas(find(GroupMat(Change_factor(j,i),:)),i) = abs((Single_OutValues(j) - Single_OutValues(j+1) )/Delt); %(2/3));
SAmeas(i,Change_factor') = abs((Single_OutValues(j) - Single_OutValues(j+1) )/Delt);
else
Change_factor(j,i) = find(Single_Sample(j+1,:)-Single_Sample(j,:));
@ -116,12 +115,17 @@ for k=1:size(Output,2)
end %for i=1:r
if NumGroups ~ 0
if NumGroups ~= 0
SAmeas = SAmeas';
end
% Compute Mu AbsMu and StDev
if any(any(isnan(SAmeas)))
AbsMu=NaN(NumFact,1);
if NumGroups == 0
Mu=NaN(NumFact,1);
StDev=NaN(NumFact,1);
end
for j=1:NumFact
SAm = SAmeas(j,:);
SAm = SAm(find(~isnan(SAm)));
@ -143,8 +147,8 @@ for k=1:size(Output,2)
% Define the output Matrix - if we have groups we cannot define the old
% measure mu, only mu* makes sense
if NumGroups > 0
OutMatrix = [OutMatrix; AbsMu];
OutMatrix((k-1)*sizea+1:k*sizea,:) = AbsMu;
else
OutMatrix = [OutMatrix; AbsMu, Mu, StDev];
OutMatrix((k-1)*sizea+1:k*sizea,:) = [AbsMu, Mu, StDev];
end
end % For Each Output

26
matlab/gsa/Sampling_Function_2.m
View File

@ -1,6 +1,7 @@
function [Outmatrix, OutFact] = Sampling_Function_2(p, k, r, UB, LB, GroupMat)
%[Outmatrix, OutFact] = Sampling_Function_2(p, k, r, UB, LB, GroupMat)
% Inputs: k (1,1) := number of factors examined or number of groups examined.
% Inputs:
% k (1,1) := number of factors examined or number of groups examined.
% In case the groups are chosen the number of factors is stores in NumFact and
% sizea becomes the number of created groups.
% NumFact (1,1) := number of factors examined in the case when groups are chosen
@ -13,7 +14,7 @@ function [Outmatrix, OutFact] = Sampling_Function_2(p, k, r, UB, LB, GroupMat)
% are set to 1 in correspondence of the factors that belong to the fixed group. All
% the other elements are zero.
% Local Variables:
% sizeb (1,1) := sizea+1
% sizeb (1,1) := sizea+1
% sizec (1,1) := 1
% randmult (sizea,1) := vector of random +1 and -1
% perm_e(1,sizea) := vector of sizea random permutated indeces
@ -34,7 +35,8 @@ function [Outmatrix, OutFact] = Sampling_Function_2(p, k, r, UB, LB, GroupMat)
% AuxMat(sizeb,sizea) := Delta*0.5*((2*B - A) * DD0 + A) in Morris, 1991. The AuxMat is used as in Morris design
% for single factor analysis, while it constitutes an intermediate step for the group analysis.
%
% Output: Outmatrix(sizeb*r, sizea) := for the entire sample size computed In(i,j) matrices
% Outputs:
% Outmatrix(sizeb*r, sizea) := for the entire sample size computed In(i,j) matrices
% OutFact(sizea*r,1) := for the entire sample size computed Fact(i,1) vectors
%
% Note: B0 is constructed as in Morris design when groups are not considered. When groups are considered the routine
@ -56,7 +58,7 @@ function [Outmatrix, OutFact] = Sampling_Function_2(p, k, r, UB, LB, GroupMat)
%
% Copyright © 2005 European Commission
% Copyright © 2012-2017 Dynare Team
% Copyright © 2012-2023 Dynare Team
%
% This file is part of Dynare.
%
@ -80,25 +82,23 @@ Delta = p/(2*p-2);
NumFact = sizea;
GroupNumber = size(GroupMat,2);
if GroupNumber ~ 0;
if GroupNumber ~= 0
sizea = size(GroupMat,2);
end
sizeb = sizea + 1;
sizec = 1;
Outmatrix = [];
OutFact = [];
% For each i generate a trajectory
for i=1:r
% Construct DD0 - OLD VERSION - it does not need communication toolbox
% RAND(N,M) is an NXM matrix with random entries, chosen from a uniform distribution on the interval (0.0,1.0).
% Note that DD0 tells if the factor have to be increased or ddecreased
% by Delta.
randmult = ones(k,1);
v = rand(k,1);
randmult (find(v < 0.5))=-1;
randmult (v < 0.5)=-1;
randmult = repmat(randmult,1,k);
DD0 = randmult .* eye(k);
@ -133,7 +133,7 @@ for i=1:r
% When groups are present the random permutation is done only on B. The effect is the same since
% the added part (A0*x0') is completely random.
if GroupNumber ~ 0
if GroupNumber ~= 0
B = B * (GroupMat*P0')';
end
@ -158,9 +158,9 @@ for i=1:r
% 3) check in which interval the random numbers fall
% 4) generate the corresponding integer
v = repmat(rand(NumFact,1),1,size(MyInt,2)+1); % 1)
IntUsed = repmat([0:1/size(MyInt,2):1],NumFact,1); % 2)
IntUsed = repmat(0:1/size(MyInt,2):1,NumFact,1); % 2)
DiffAuxVec = IntUsed - v; % 3)
w=NaN(1,size(DiffAuxVec,1));
for ii = 1:size(DiffAuxVec,1)
w(1,ii) = max(find(DiffAuxVec(ii,:)<0)); % 4)
end
@ -168,7 +168,7 @@ for i=1:r
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% b --> Compute the matrix B*, here indicated as B0. Each row in B0 is a
% trajectory for Morris Calculations. The dimension of B0 is (Numfactors+1,Numfactors)
if GroupNumber ~ 0
if GroupNumber ~= 0
B0 = (A0*x0' + AuxMat);
else
B0 = (A0*x0' + AuxMat)*P0;
@ -183,6 +183,7 @@ for i=1:r
% Create the Factor vector. Each component of this vector indicate which factor or group of factor
% has been changed in each step of the trajectory.
Fact=NaN(1,sizea);
for j=1:sizea
Fact(1,j) = find(P0(j,:));
end
@ -190,5 +191,4 @@ for i=1:r
Outmatrix = [Outmatrix; In];
OutFact = [OutFact; Fact'];
end

13
matlab/gsa/cumplot.m
View File

@ -1,5 +1,10 @@
function h = cumplot(x)
%function h =cumplot(x)
% Inputs:
% - x [double] data series
%
% Outputs:
% - h [handle] figure handle
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
@ -26,9 +31,5 @@ function h = cumplot(x)
n=length(x);
x=[-inf; sort(x); Inf];
y=[0:n n]./n;
h0 = stairs(x,y);
grid on,
if nargout
h=h0;
end
h = stairs(x,y);
grid on

287
matlab/gsa/filt_mc_.m
View File

@ -1,5 +1,25 @@
function [rmse_MC, ixx] = filt_mc_(OutDir,options_gsa_,dataset_,dataset_info)
% function [rmse_MC, ixx] = filt_mc_(OutDir)
function [rmse_MC, ixx] = filt_mc_(OutDir,options_gsa_,dataset_,dataset_info,M_,oo_,options_,bayestopt_,estim_params_)
% [rmse_MC, ixx] = filt_mc_(OutDir,options_gsa_,dataset_,dataset_info,M_,oo_,options_,bayestopt_,estim_params_
% Inputs:
% - OutputDirectoryName [string] name of the output directory
% - options_gsa_ [structure] GSA options
% - dataset_ [dseries] object storing the dataset
% - dataset_info [structure] storing informations about the sample.
% - M_ [structure] Matlab's structure describing the model
% - oo_ [structure] storing the results
% - options_ [structure] Matlab's structure describing the current options
% - bayestopt_ [structure] describing the priors
% - estim_params_ [structure] characterizing parameters to be estimated
%
% Outputs:
% - rmse_MC [double] RMSE by nvar matrix of the RMSEs
% - ixx [double] RMSE by nvar matrix of sorting
% indices (descending order of RMSEs)
%
% Notes: the R^2 definition is 1-var(ymodel-ydata)/var(ydata). It ranges
% bewteen (-inf, 1], with negative values indicating that themodel is a worse
% predictor than the sample mean of the data
% inputs (from opt_gsa structure)
% vvarvecm = options_gsa_.var_rmse;
% loadSA = options_gsa_.load_rmse;
@ -7,7 +27,6 @@ function [rmse_MC, ixx] = filt_mc_(OutDir,options_gsa_,dataset_,dataset_info)
% alpha = options_gsa_.alpha_rmse;
% alpha2 = options_gsa_.alpha2_rmse;
% istart = options_gsa_.istart_rmse;
% alphaPC = 0.5;
%
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
@ -31,9 +50,6 @@ function [rmse_MC, ixx] = filt_mc_(OutDir,options_gsa_,dataset_,dataset_info)
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
global bayestopt_ estim_params_ M_ options_ oo_
% options_gsa_=options_.opt_gsa;
vvarvecm = options_gsa_.var_rmse;
if options_.TeX
vvarvecm_tex = options_gsa_.var_rmse_tex;
@ -46,13 +62,10 @@ alpha = options_gsa_.alpha_rmse;
alpha2 = 0;
pvalue = options_gsa_.alpha2_rmse;
istart = max(2,options_gsa_.istart_rmse);
alphaPC = 0.5;
fname_ = M_.fname;
lgy_ = M_.endo_names;
dr_ = oo_.dr;
skipline(2)
skipline(1)
disp('Starting sensitivity analysis')
disp('for the fit of EACH observed series ...')
skipline()
@ -61,12 +74,12 @@ if ~options_.nograph
a=dir([OutDir,filesep,'*.*']);
tmp1='0';
if options_.opt_gsa.ppost
tmp=['_rmse_post'];
tmp='_rmse_post';
else
if options_.opt_gsa.pprior
tmp=['_rmse_prior'];
tmp='_rmse_prior';
else
tmp=['_rmse_mc'];
tmp='_rmse_mc';
end
if options_gsa_.lik_only
tmp1 = [tmp,'_post_SA'];
@ -75,17 +88,23 @@ if ~options_.nograph
end
for j=1:length(a)
if strmatch([fname_,tmp],a(j).name)
disp(a(j).name)
if options_.debug
disp(a(j).name)
end
delete([OutDir,filesep,a(j).name])
end
if strmatch([fname_,tmp1],a(j).name)
disp(a(j).name)
if options_.debug
disp(a(j).name)
end
delete([OutDir,filesep,a(j).name])
end
end
disp('done !')
end
[param_names,param_names_tex]=get_LaTeX_parameter_names(M_,options_,estim_params_,bayestopt_);
nshock=estim_params_.nvx + estim_params_.nvn + estim_params_.ncx + estim_params_.ncn;
npar=estim_params_.np;
if ~isempty(options_.mode_file)
@ -94,10 +113,12 @@ end
if options_.opt_gsa.ppost
c=load([M_.dname filesep 'Output' filesep fname_,'_mean.mat'],'xparam1');
xparam1_mean=c.xparam1;
xparam1=c.xparam1;
clear c
elseif ~isempty(options_.mode_file) && exist([M_.dname filesep 'Output' filesep fname_,'_mean.mat'])==2
elseif ~isempty(options_.mode_file) && exist([M_.dname filesep 'Output' filesep fname_,'_mean.mat'],'file')==2
c=load([M_.dname filesep 'Output' filesep fname_,'_mean.mat'],'xparam1');
xparam1_mean=c.xparam1;
xparam1=c.xparam1;
clear c
end
@ -124,31 +145,11 @@ if loadSA
end
end
if ~loadSA
if exist('xparam1','var')
M_ = set_all_parameters(xparam1,estim_params_,M_);
ys_mode=evaluate_steady_state(oo_.steady_state,[oo_.exo_steady_state; oo_.exo_det_steady_state],M_,options_,~options_.steadystate.nocheck);
end
if exist('xparam1_mean','var')
M_ = set_all_parameters(xparam1_mean,estim_params_,M_);
ys_mean=evaluate_steady_state(oo_.steady_state,[oo_.exo_steady_state; oo_.exo_det_steady_state],M_,options_,~options_.steadystate.nocheck);
end
Y = transpose(dataset_.data);
gend = dataset_.nobs;
data_index = dataset_info.missing.aindex;
missing_value = dataset_info.missing.state;
for jx=1:gend
data_indx(jx,data_index{jx})=true;
end
load([DirectoryName filesep M_.fname '_data.mat']);
filfilt = dir([DirectoryName filesep M_.fname '_filter_step_ahead*.mat']);
temp_smooth_file_list = dir([DirectoryName filesep M_.fname '_smooth*.mat']);
jfile=0;
for j=1:length(temp_smooth_file_list)
if isempty(strfind(temp_smooth_file_list(j).name,'smoothed')),
jfile=jfile+1;
filsmooth(jfile)=temp_smooth_file_list(j);
end
end
filupdate = dir([DirectoryName filesep M_.fname '_update*.mat']);
filparam = dir([DirectoryName filesep M_.fname '_param*.mat']);
x=[];
@ -156,11 +157,11 @@ if ~loadSA
sto_ys=[];
for j=1:length(filparam)
if isempty(strmatch([M_.fname '_param_irf'],filparam(j).name))
load([DirectoryName filesep filparam(j).name]);
x=[x; stock];
logpo2=[logpo2; stock_logpo];
sto_ys=[sto_ys; stock_ys];
clear stock stock_logpo stock_ys;
temp=load([DirectoryName filesep filparam(j).name]); % from prior_posterior_statistics_core
x=[x; temp.stock];
logpo2=[logpo2; temp.stock_logpo];
sto_ys=[sto_ys; temp.stock_ys];
clear temp;
end
end
nruns=size(x,1);
@ -168,38 +169,41 @@ if ~loadSA
if options_.opt_gsa.ppost || (options_.opt_gsa.ppost==0 && options_.opt_gsa.lik_only==0)
skipline()
disp('Computing RMSE''s...')
jxj=NaN(length(vvarvecm),1);
js=NaN(length(vvarvecm),1);
yss=NaN(length(vvarvecm),gend,size(sto_ys,1));
for i = 1:length(vvarvecm)
vj = vvarvecm{i};
jxj(i) = strmatch(vj, lgy_(dr_.order_var), 'exact');
js(i) = strmatch(vj, lgy_, 'exact');
jxj(i) = strmatch(vj, M_.endo_names(oo_.dr.order_var), 'exact');
js(i) = strmatch(vj, M_.endo_names, 'exact');
yss(i,:,:)=repmat(sto_ys(:,js(i))',[gend,1]);
end
if exist('xparam1','var')
[alphahat,etahat,epsilonhat,ahat,SteadyState,trend_coeff,aK] = DsgeSmoother(xparam1,gend,Y,data_index,missing_value,M_,oo_,options_,bayestopt_,estim_params_);
y0 = reshape( squeeze(aK(1,jxj,1:gend)),[gend length(jxj)]);% + kron(ys_mode(js),ones(1,gend)));
yobs = transpose( ahat(jxj,:));% + kron(ys_mode(js),ones(1,gend)));
[~,~,~,ahat,~,~,aK] = DsgeSmoother(xparam1,gend,Y,data_index,missing_value,M_,oo_,options_,bayestopt_,estim_params_);
y0 = reshape( squeeze(aK(1,jxj,1:gend)),[gend length(jxj)]);
yobs = transpose( ahat(jxj,:));
rmse_mode = sqrt(mean((yobs(istart:end,:)-y0(istart:end,:)).^2));
r2_mode = 1-sum((yobs(istart:end,:)-y0(istart:end,:)).^2)./sum(yobs(istart:end,:).^2);
end
y0=-yss;
y0=-yss; %demean everything using the theoretical mean, i.e. steady state
nbb=0;
for j=1:length(filfilt)
load([DirectoryName filesep M_.fname '_filter_step_ahead',num2str(j),'.mat']);
nb = size(stock,4);
y0(:,:,nbb+1:nbb+nb)=y0(:,:,nbb+1:nbb+nb)+reshape(stock(1,js,1:gend,:),[length(js) gend nb]);
temp=load([DirectoryName filesep M_.fname '_filter_step_ahead',num2str(j),'.mat']);
nb = size(temp.stock,4);
y0(:,:,nbb+1:nbb+nb)=y0(:,:,nbb+1:nbb+nb)+reshape(temp.stock(1,js,1:gend,:),[length(js) gend nb]);
nbb=nbb+nb;
clear stock;
clear temp;
end
yobs=-yss;
nbb=0;
for j=1:length(filupdate)
load([DirectoryName filesep M_.fname '_update',num2str(j),'.mat']);
nb = size(stock,3);
yobs(:,:,nbb+1:nbb+nb)=yobs(:,:,nbb+1:nbb+nb)+reshape(stock(js,1:gend,:),[length(js) gend nb]);
temp=load([DirectoryName filesep M_.fname '_update',num2str(j),'.mat']);
nb = size(temp.stock,3);
yobs(:,:,nbb+1:nbb+nb)=yobs(:,:,nbb+1:nbb+nb)+reshape(temp.stock(js,1:gend,:),[length(js) gend nb]);
nbb=nbb+nb;
clear stock;
clear temp;
end
y0M=mean(y0,2);
rmse_MC=zeros(nruns,length(js));
r2_MC=zeros(nruns,length(js));
for j=1:nruns
@ -207,14 +211,15 @@ if ~loadSA
r2_MC(j,:) = 1-mean((yobs(:,istart:end,j)'-y0(:,istart:end,j)').^2)./mean((yobs(:,istart:end,j)').^2);
end
if exist('xparam1_mean','var')
[alphahat,etahat,epsilonhat,ahat,SteadyState,trend_coeff,aK] = DsgeSmoother(xparam1_mean,gend,Y,data_index,missing_value,M_,oo_,options_,bayestopt_,estim_params_);
y0 = reshape( squeeze(aK(1,jxj,1:gend)),[gend length(jxj)]);% + kron(ys_mean(js),ones(1,gend)));
yobs = transpose( ahat(jxj,:));% + kron(ys_mean(js),ones(1,gend)));
[~,~,~,ahat,~,~,aK] = DsgeSmoother(xparam1_mean,gend,Y,data_index,missing_value,M_,oo_,options_,bayestopt_,estim_params_);
y0 = reshape( squeeze(aK(1,jxj,1:gend)),[gend length(jxj)]);
yobs = transpose( ahat(jxj,:));
rmse_pmean = sqrt(mean((yobs(istart:end,:)-y0(istart:end,:)).^2));
r2_pmean = 1-mean((yobs(istart:end,:)-y0(istart:end,:)).^2)./mean(yobs(istart:end,:).^2);
end
clear stock_filter;
end
lnprior=NaN(nruns,1);
for j=1:nruns
lnprior(j,1) = priordens(x(j,:)',bayestopt_.pshape,bayestopt_.p6,bayestopt_.p7,bayestopt_.p3,bayestopt_.p4);
end
@ -242,7 +247,7 @@ if ~loadSA
end
end
end
else
else % loadSA
if options_.opt_gsa.lik_only && options_.opt_gsa.ppost==0
load([OutDir,filesep,fnamtmp, '.mat'],'x','logpo2','likelihood');
else
@ -252,27 +257,27 @@ else
nruns=size(x,1);
nfilt=floor(pfilt*nruns);
end
% smirnov tests
% Smirnov tests
nfilt0 = nfilt*ones(length(vvarvecm), 1);
logpo2=logpo2(:);
if ~options_.opt_gsa.ppost
[dum, ipost]=sort(-logpo2);
[dum, ilik]=sort(-likelihood);
[~, ipost]=sort(-logpo2);
[~, ilik]=sort(-likelihood);
end
% visual scatter analysis!
if options_.opt_gsa.ppost
tmp_title='R2 Posterior:';
atitle='R2 Posterior:';
tmp_title='R2 Scatter plot: Posterior';
atitle='R2 Scatter plot: Posterior';
asname='r2_post';
else
if options_.opt_gsa.pprior
tmp_title='R2 Prior:';
atitle='R2 Prior:';
tmp_title='R2 Scatter plot: Prior';
atitle='R2 Scatter plot: Prior';
asname='r2_prior';
else
tmp_title='R2 MC:';
atitle='R2 MC:';
tmp_title='R2 Scatter plot: MC';
atitle='R2 Scatter plot: MC';
asname='r2_mc';
end
end
@ -297,13 +302,10 @@ if ~options_.opt_gsa.ppost && options_.opt_gsa.lik_only
options_mcf.pvalue_ks = alpha;
options_mcf.pvalue_corr = pvalue;
options_mcf.alpha2 = alpha2;
options_mcf.param_names = param_names;
if options_.TeX
[pnames,pnames_tex]=get_LaTeX_parameter_names(M_,options_,estim_params_,bayestopt_);
options_mcf.param_names = pnames;
options_mcf.param_names_tex = pnames_tex;
options_mcf.param_names_tex = param_names_tex;
else
[pnames]=get_LaTeX_parameter_names(M_,options_,estim_params_,bayestopt_);
options_mcf.param_names = pnames;
options_mcf.param_names_tex = {};
end
options_mcf.fname_ = fname_;
@ -313,7 +315,12 @@ if ~options_.opt_gsa.ppost && options_.opt_gsa.lik_only
options_mcf.title = atitle;
options_mcf.beha_title = 'better posterior kernel';
options_mcf.nobeha_title = 'worse posterior kernel';
mcf_analysis(x, ipost(1:nfilt), ipost(nfilt+1:end), options_mcf, options_);
if options_.TeX
options_mcf.beha_title_latex = 'better posterior kernel';
options_mcf.nobeha_title_latex = 'worse posterior kernel';
end
mcf_analysis(x, ipost(1:nfilt), ipost(nfilt+1:end), options_mcf, M_, options_, bayestopt_, estim_params_);
if options_.opt_gsa.pprior
anam = 'rmse_prior_lik';
atitle = 'RMSE prior: Log Likelihood Kernel';
@ -326,15 +333,20 @@ if ~options_.opt_gsa.ppost && options_.opt_gsa.lik_only
options_mcf.title = atitle;
options_mcf.beha_title = 'better likelihood';
options_mcf.nobeha_title = 'worse likelihood';
mcf_analysis(x, ilik(1:nfilt), ilik(nfilt+1:end), options_mcf, options_);
if options_.TeX
options_mcf.beha_title_latex = 'better likelihood';
options_mcf.nobeha_title_latex = 'worse likelihood';
end
mcf_analysis(x, ilik(1:nfilt), ilik(nfilt+1:end), options_mcf, M_, options_, bayestopt_, estim_params_);
else
if options_.opt_gsa.ppost
rmse_txt=rmse_pmean;
r2_txt=r2_pmean;
else
if options_.opt_gsa.pprior || ~exist('rmse_pmean')
if exist('rmse_mode')
if options_.opt_gsa.pprior || ~exist('rmse_pmean','var')
if exist('rmse_mode','var')
rmse_txt=rmse_mode;
r2_txt=r2_mode;
else
@ -346,18 +358,19 @@ else
r2_txt=r2_pmean;
end
end
ixx=NaN(size(rmse_MC,1),length(vvarvecm));
for i = 1:length(vvarvecm)
[dum, ixx(:,i)] = sort(rmse_MC(:,i));
[~, ixx(:,i)] = sort(rmse_MC(:,i));
end
PP = ones(npar+nshock, length(vvarvecm));
PPV = ones(length(vvarvecm), length(vvarvecm), npar+nshock);
SS = zeros(npar+nshock, length(vvarvecm));
for j = 1:npar+nshock
for i = 1:length(vvarvecm)
[H, P, KSSTAT] = smirnov(x(ixx(nfilt0(i)+1:end,i),j),x(ixx(1:nfilt0(i),i),j), alpha);
[H1, P1, KSSTAT1] = smirnov(x(ixx(nfilt0(i)+1:end,i),j),x(ixx(1:nfilt0(i),i),j),alpha,1);
[H2, P2, KSSTAT2] = smirnov(x(ixx(nfilt0(i)+1:end,i),j),x(ixx(1:nfilt0(i),i),j),alpha,-1);
if H1 & H2==0
[~, P] = smirnov(x(ixx(nfilt0(i)+1:end,i),j),x(ixx(1:nfilt0(i),i),j), alpha);
[H1] = smirnov(x(ixx(nfilt0(i)+1:end,i),j),x(ixx(1:nfilt0(i),i),j),alpha,1);
[H2] = smirnov(x(ixx(nfilt0(i)+1:end,i),j),x(ixx(1:nfilt0(i),i),j),alpha,-1);
if H1==0 && H2==0
SS(j,i)=1;
elseif H1==0
SS(j,i)=-1;
@ -369,7 +382,7 @@ else
for i = 1:length(vvarvecm)
for l = 1:length(vvarvecm)
if l~=i && PP(j,i)<alpha && PP(j,l)<alpha
[H,P,KSSTAT] = smirnov(x(ixx(1:nfilt0(i),i),j),x(ixx(1:nfilt0(l),l),j), alpha);
[~,P] = smirnov(x(ixx(1:nfilt0(i),i),j),x(ixx(1:nfilt0(l),l),j), alpha);
PPV(i,l,j) = P;
elseif l==i
PPV(i,l,j) = PP(j,i);
@ -400,7 +413,11 @@ else
set(h,'color','k','linewidth',1)
h=cumplot(lnprior(ixx(nfilt0(i)+1:end,i)));
set(h,'color','red','linewidth',2)
title(vvarvecm{i},'interpreter','none')
if options_.TeX
title(vvarvecm_tex{i},'interpreter','latex')
else
title(vvarvecm{i},'interpreter','none')
end
if mod(i,9)==0 || i==length(vvarvecm)
if ~isoctave
annotation('textbox', [0.1,0,0.35,0.05],'String', 'Log-prior for BETTER R2','Color','Blue','horizontalalignment','center');
@ -448,7 +465,11 @@ else
set(h,'color','k','linewidth',1)
h=cumplot(likelihood(ixx(nfilt0(i)+1:end,i)));
set(h,'color','red','linewidth',2)
title(vvarvecm{i},'interpreter','none')
if options_.TeX
title(vvarvecm_tex{i},'interpreter','latex')
else
title(vvarvecm{i},'interpreter','none')
end
if options_.opt_gsa.ppost==0
set(gca,'xlim',[min( likelihood(ixx(1:nfilt0(i),i)) ) max( likelihood(ixx(1:nfilt0(i),i)) )])
end
@ -499,7 +520,11 @@ else
set(h,'color','k','linewidth',1)
h=cumplot(logpo2(ixx(nfilt0(i)+1:end,i)));
set(h,'color','red','linewidth',2)
title(vvarvecm{i},'interpreter','none')
if options_.TeX
title(vvarvecm_tex{i},'interpreter','latex')
else
title(vvarvecm{i},'interpreter','none')
end
if options_.opt_gsa.ppost==0
set(gca,'xlim',[min( logpo2(ixx(1:nfilt0(i),i)) ) max( logpo2(ixx(1:nfilt0(i),i)) )])
end
@ -529,15 +554,6 @@ else
end
end
end
if options_.TeX
[pnames,pnames_tex]=get_LaTeX_parameter_names(M_,options_,estim_params_,bayestopt_);
param_names = pnames;
param_names_tex = pnames_tex;
else
[pnames]=get_LaTeX_parameter_names(M_,options_,estim_params_,bayestopt_);
param_names = pnames;
param_names_tex = {};
end
skipline()
title_string='RMSE over the MC sample:';
data_mat=[min(rmse_MC)' max(rmse_MC)'];
@ -549,7 +565,7 @@ else
end
invar = find( std(rmse_MC)./mean(rmse_MC)<=0.0001 );
if ~isempty(invar)
skipline(2)
skipline(1)
disp('RMSE is not varying significantly over the MC sample for the following variables:')
disp(vvarvecm{invar})
disp('These variables are excluded from SA')
@ -561,8 +577,7 @@ else
rmse_MC = rmse_MC(:,ivar);
skipline()
disp(['Sample filtered the ',num2str(pfilt*100),'% best RMSE''s for each observed series ...' ])
skipline(2)
disp('RMSE ranges after filtering:')
skipline(1)
title_string='RMSE ranges after filtering:';
if options_.opt_gsa.ppost==0 && options_.opt_gsa.pprior
headers = {'Variable'; 'min'; 'max'; 'min'; 'max'; 'posterior mode'};
@ -589,7 +604,7 @@ else
else
values_length = max(ceil(max(max(log10(abs(data_mat(isfinite(data_mat))))))),1)+val_precis+1;
end
if any(data_mat) < 0 %add one character for minus sign
if any(data_mat < 0) %add one character for minus sign
values_length = values_length+1;
end
headers_length = cellofchararraymaxlength(headers(2:end));
@ -598,7 +613,6 @@ else
else
val_width = max(headers_length, values_length)+2;
end
value_format = sprintf('%%%d.%df',val_width,val_precis);
header_string_format = sprintf('%%%ds',val_width);
if options_.opt_gsa.ppost==0 && options_.opt_gsa.pprior
optional_header=sprintf([label_format_leftbound,header_string_format,header_string_format,header_string_format,header_string_format],'','',['best ',num2str(pfilt*100),'% filtered'],'','remaining 90%');
@ -610,7 +624,7 @@ else
if options_.opt_gsa.ppost==0 && options_.opt_gsa.pprior
optional_header={[' & \multicolumn{2}{c}{best ',num2str(pfilt*100),' filtered} & \multicolumn{2}{c}{remaining 90\%}\\']};
else
optional_header={[' & \multicolumn{2}{c}{best filtered} & \multicolumn{2}{c}{remaining}\\']};
optional_header={' & \multicolumn{2}{c}{best filtered} & \multicolumn{2}{c}{remaining}\\'};
end
dyn_latex_table(M_, options_, title_string, 'RMSE_ranges_after_filtering', headers_tex, vvarvecm_tex, data_mat, 0, val_width, val_precis, optional_header);
end
@ -657,7 +671,7 @@ else
else
values_length = max(ceil(max(max(log10(abs(data_mat(isfinite(data_mat))))))),1)+val_precis+1;
end
if any(data_mat) < 0 %add one character for minus sign
if any(data_mat < 0) %add one character for minus sign
values_length = values_length+1;
end
headers_length = cellofchararraymaxlength(headers(2:end));
@ -666,7 +680,6 @@ else
else
val_width = max(headers_length, values_length)+2;
end
value_format = sprintf('%%%d.%df',val_width,val_precis);
header_string_format = sprintf('%%%ds',val_width);
if options_.opt_gsa.ppost==0 && options_.opt_gsa.pprior
@ -679,7 +692,7 @@ else
if ~options_.opt_gsa.ppost && options_.opt_gsa.pprior
optional_header = {[' & \multicolumn{2}{c}{best ',num2str(pfilt*100),' filtered} & \multicolumn{2}{c}{remaining 90\%}\\']};
else
optional_header = {[' & \multicolumn{2}{c}{best filtered} & \multicolumn{2}{c}{remaining}\\']};
optional_header = {' & \multicolumn{2}{c}{best filtered} & \multicolumn{2}{c}{remaining}\\'};
end
dyn_latex_table(M_, options_, title_string, 'R2_ranges_after_filtering', headers_tex, vvarvecm_tex, data_mat, 0, val_width, val_precis, optional_header);
end
@ -690,16 +703,15 @@ else
SP(ns,j)=ones(size(ns));
SS(:,j)=SS(:,j).*SP(:,j);
end
for j=1:npar+nshock %estim_params_.np,
nsp=NaN(npar+nshock,1);
for j=1:npar+nshock
nsp(j)=length(find(SP(j,:)));
end
snam0=param_names(find(nsp==0));
snam1=param_names(find(nsp==1));
snam2=param_names(find(nsp>1));
snam=param_names(find(nsp>0));
snam0=param_names(nsp==0);
snam1=param_names(nsp==1);
snam2=param_names(nsp>1);
nsnam=(find(nsp>1));
skipline(2)
skipline(1)
disp('These parameters do not affect significantly the fit of ANY observed series:')
disp(char(snam0))
skipline()
@ -708,7 +720,6 @@ else
skipline()
disp('These parameters affect MORE THAN ONE observed series: trade off exists!')
disp(char(snam2))
pnam=bayestopt_.name;
% plot trade-offs
if ~options_.nograph
a00=jet(length(vvarvecm));
@ -740,8 +751,12 @@ else
options_mcf.amcf_title = [atitle ' ' vvarvecm{iy}];
options_mcf.beha_title = ['better fit of ' vvarvecm{iy}];
options_mcf.nobeha_title = ['worse fit of ' vvarvecm{iy}];
if options_.TeX
options_mcf.beha_title_latex = ['better fit of ' vvarvecm_tex{iy}];
options_mcf.nobeha_title_latex = ['worse fit of ' vvarvecm_tex{iy}];
end
options_mcf.title = ['the fit of ' vvarvecm{iy}];
mcf_analysis(x, ixx(1:nfilt0(iy),iy), ixx(nfilt0(iy)+1:end,iy), options_mcf, options_);
mcf_analysis(x, ixx(1:nfilt0(iy),iy), ixx(nfilt0(iy)+1:end,iy), options_mcf, M_, options_, bayestopt_, estim_params_);
end
for iy = 1:length(vvarvecm)
ipar = find(any(squeeze(PPV(iy,:,:))<alpha));
@ -767,35 +782,43 @@ else
set(h0,'color',a00(i,:),'linewidth',2)
end
ydum=get(gca,'ylim');
if exist('xparam1')
if exist('xparam1','var')
xdum=xparam1(ipar(j));
h1=plot([xdum xdum],ydum);
set(h1,'color',[0.85 0.85 0.85],'linewidth',2)
end
xlabel('')
title([pnam{ipar(j)}],'interpreter','none')
if options_.TeX
title([param_names_tex{ipar(j)}],'interpreter','latex')
else
title([param_names{ipar(j)}],'interpreter','none')
end
end
if isoctave
legend(vertcat('base',vvarvecm),'location','eastoutside');
else
h0=legend(vertcat('base',vvarvecm));
set(h0,'fontsize',6,'position',[0.7 0.1 0.2 0.3],'interpreter','none');
if options_.TeX
h0=legend(vertcat('base',vvarvecm_tex),'interpreter','latex');
else
h0=legend(vertcat('base',vvarvecm),'interpreter','none');
end
set(h0,'fontsize',6,'position',[0.7 0.1 0.2 0.3]);
end
if options_.opt_gsa.ppost
dyn_saveas(hh_fig,[ OutDir filesep fname_ '_rmse_post_' vvarvecm{iy} '_' int2str(ix)],options_.nodisplay,options_.graph_format);
if options_.TeX
create_TeX_loader(options_,[ OutDir filesep fname_ '_rmse_post_' vvarvecm{iy} '_' int2str(ix)],ix,[temp_name,' observed variable $',vvarvecm_tex{iy} '$'],['rmse_post_' vvarvecm{iy}],1)
create_TeX_loader(options_,[ OutDir filesep fname_ '_rmse_post_' vvarvecm{iy} '_' int2str(ix)],ix,[temp_name,' observed variable ',vvarvecm_tex{iy} ],['rmse_post_' vvarvecm{iy}],1)
end
else
if options_.opt_gsa.pprior
dyn_saveas(hh_fig,[OutDir filesep fname_ '_rmse_prior_' vvarvecm{iy} '_' int2str(ix) ],options_.nodisplay,options_.graph_format);
if options_.TeX
create_TeX_loader(options_,[OutDir filesep fname_ '_rmse_prior_' vvarvecm{iy} '_' int2str(ix) ],ix,[temp_name,' observed variable $',vvarvecm_tex{iy} '$'],['rmse_prior_' vvarvecm{iy}],1)
create_TeX_loader(options_,[OutDir filesep fname_ '_rmse_prior_' vvarvecm{iy} '_' int2str(ix) ],ix,[temp_name,' observed variable ',vvarvecm_tex{iy}],['rmse_prior_' vvarvecm{iy}],1)
end
else
dyn_saveas(hh_fig,[OutDir filesep fname_ '_rmse_mc_' vvarvecm{iy} '_' int2str(ix)],options_.nodisplay,options_.graph_format);
if options_.TeX
create_TeX_loader(options_,[OutDir filesep fname_ '_rmse_mc_' vvarvecm{iy} '_' int2str(ix)],ix,[temp_name,' observed variable $',vvarvecm_tex{iy} '$'],['rmse_mc_' vvarvecm{iy}],1)
create_TeX_loader(options_,[OutDir filesep fname_ '_rmse_mc_' vvarvecm{iy} '_' int2str(ix)],ix,[temp_name,' observed variable ',vvarvecm_tex{iy}],['rmse_mc_' vvarvecm{iy}],1)
end
end
end
@ -824,20 +847,28 @@ else
set(h0,'color',a00(i,:),'linewidth',2)
end
ydum=get(gca,'ylim');
if exist('xparam1')
if exist('xparam1','var')
xdum=xparam1(nsnam(j));
h1=plot([xdum xdum],ydum);
set(h1,'color',[0.85 0.85 0.85],'linewidth',2)
end
xlabel('')
title([pnam{nsnam(j)}],'interpreter','none')
if options_.TeX
title([param_names_tex{nsnam(j)}],'interpreter','latex')
else
title([param_names{nsnam(j)}],'interpreter','none')
end
end
%subplot(3,2,6)
if isoctave
legend(vertcat('base',vvarvecm),'location','eastoutside');
else
h0=legend(vertcat('base',vvarvecm));
set(h0,'fontsize',6,'position',[0.7 0.1 0.2 0.3],'interpreter','none');
if options_.TeX
h0=legend(vertcat('base',vvarvecm_tex),'interpreter','latex');
else
h0=legend(vertcat('base',vvarvecm),'interpreter','none');
end
set(h0,'fontsize',6,'position',[0.7 0.1 0.2 0.3]);
end
if options_.opt_gsa.ppost
dyn_saveas(hh_fig,[ OutDir filesep fname_ '_rmse_post_params_' int2str(ix)],options_.nodisplay,options_.graph_format);
@ -885,11 +916,11 @@ pnames=cell(np,1);
pnames_tex=cell(np,1);
for ii=1:length(bayestopt_.name)
if options_.TeX
[param_name_temp, param_name_tex_temp]= get_the_name(ii,options_.TeX,M_,estim_params_,options_);
pnames_tex{ii,1} = strrep(param_name_tex_temp,'$','');
[param_name_temp, param_name_tex_temp]= get_the_name(ii,options_.TeX,M_,estim_params_,options_.varobs);
pnames_tex{ii,1} = param_name_tex_temp;
pnames{ii,1} = param_name_temp;
else
param_name_temp = get_the_name(ii,options_.TeX,M_,estim_params_,options_);
param_name_temp = get_the_name(ii,options_.TeX,M_,estim_params_,options_.varobs);
pnames{ii,1} = param_name_temp;
end
end

99
matlab/gsa/gsa_plotmatrix.m
View File

@ -1,99 +0,0 @@
function gsa_plotmatrix(type,varargin)
% function gsa_plotmatrix(type,varargin)
% extended version of the standard MATLAB plotmatrix
%
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
% marco.ratto@ec.europa.eu
% Copyright © 2011-2012 European Commission
% Copyright © 2011-2017 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 <https://www.gnu.org/licenses/>.
global bayestopt_ options_ M_
RootDirectoryName = CheckPath('gsa',M_.dname);
if options_.opt_gsa.pprior
load([ RootDirectoryName filesep M_.fname '_prior.mat'],'lpmat0','lpmat','istable','iunstable','iindeterm','iwrong')
else
load([ RootDirectoryName filesep M_.fname '_mc.mat'],'lpmat0','lpmat','istable','iunstable','iindeterm','iwrong')
eval(['load ' options_.mode_file ' xparam1;']');
end
iexplosive = iunstable(~ismember(iunstable,[iindeterm;iwrong]));
switch type
case 'all'
x=[lpmat0 lpmat];
NumberOfDraws=size(x,1);
B=NumberOfDraws;
case 'stable'
x=[lpmat0(istable,:) lpmat(istable,:)];
NumberOfDraws=size(x,1);
B=NumberOfDraws;
case 'nosolution'
x=[lpmat0(iunstable,:) lpmat(iunstable,:)];
NumberOfDraws=size(x,1);
B=NumberOfDraws;
case 'unstable'
x=[lpmat0(iexplosive,:) lpmat(iexplosive,:)];
NumberOfDraws=size(x,1);
B=NumberOfDraws;
case 'indeterm'
x=[lpmat0(iindeterm,:) lpmat(iindeterm,:)];
NumberOfDraws=size(x,1);
B=NumberOfDraws;
case 'wrong'
x=[lpmat0(iwrong,:) lpmat(iwrong,:)];
NumberOfDraws=size(x,1);
B=NumberOfDraws;
end
if isempty(x)
disp('Empty parameter set!')
return
end
for j=1:length(varargin)
jcol(j)=strmatch(varargin{j},bayestopt_.name,'exact');
end
[H,AX,BigA,P,PAx]=plotmatrix(x(:,jcol));
for j=1:length(varargin)
% axes(AX(1,j)), title(varargin{j})
% axes(AX(j,1)), ylabel(varargin{j})
% set(AX(1,j),'title',varargin{j}),
set(get(AX(j,1),'ylabel'),'string',varargin{j})
set(get(AX(end,j),'xlabel'),'string',varargin{j})
end
if options_.opt_gsa.pprior==0
xparam1=xparam1(jcol);
for j=1:length(varargin)
for i=1:j-1
axes(AX(j,i))
hold on, plot(xparam1(i),xparam1(j),'*r')
end
for i=j+1:length(varargin)
axes(AX(j,i))
hold on, plot(xparam1(i),xparam1(j),'*r')
end
end
end

10
matlab/gsa/gsa_skewness.m
View File

@ -1,11 +1,17 @@
function s=gsa_skewness(y)
% s=gsa_skewness(y)
% Compute normalized skewness of y
% Inputs:
% - y [double] input vector
% Outputs:
% - s [double] standardized skewness
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
% marco.ratto@ec.europa.eu
% Copyright © 2012 European Commission
% Copyright © 2012-2017 Dynare Team
% Copyright © 2012-2023 Dynare Team
%
% This file is part of Dynare.
%
@ -22,8 +28,6 @@ function s=gsa_skewness(y)
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
% y=stand_(y);
% s=mean(y.^3);
m2=mean((y-mean(y)).^2);
m3=mean((y-mean(y)).^3);
s=m3/m2^1.5;

58
matlab/gsa/gsa_speed.m
View File

@ -1,58 +0,0 @@
function [tadj, iff] = gsa_speed(A,B,mf,p)
% [tadj, iff] = gsa_speed(A,B,mf,p),
%
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
% marco.ratto@ec.europa.eu
% Copyright © 2012 European Commission
% Copyright © 2012-2017 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 <https://www.gnu.org/licenses/>.
nvar=length(mf);
nstate= size(A,1);
nshock = size(B,2);
nrun=size(B,3);
iff=zeros(nvar,nshock,nrun);
tadj=iff;
disp('Computing speed of adjustement ...')
h = dyn_waitbar(0,'Speed of adjustement...');
for i=1:nrun
irf=zeros(nvar,nshock);
a=squeeze(A(:,:,i));
b=squeeze(B(:,:,i));
IFF=inv(eye(nstate)-a)*b;
iff(:,:,i)=IFF(mf,:);
IF=IFF-b;
t=0;
while any(any(irf<0.5))
t=t+1;
IFT=((eye(nstate)-a^(t+1))*inv(eye(nstate)-a))*b-b;
irf=IFT(mf,:)./(IF(mf,:)+eps);
irf = irf.*(abs(IF(mf,:))>1.e-7)+(abs(IF(mf,:))<=1.e-7);
%irf=ft(mf,:);
tt=(irf>0.5).*t;
tadj(:,:,i)=((tt-tadj(:,:,i))==tt).*tt+tadj(:,:,i);
end
dyn_waitbar(i/nrun,h)
end
skipline()
disp('.. done !')
dyn_waitbar_close(h)

41
matlab/gsa/log_trans_.m
View File

@ -1,4 +1,29 @@
function [yy, xdir, isig, lam]=log_trans_(y0,xdir0,isig,lam)
% [yy, xdir, isig, lam]=log_trans_(y0,xdir0,isig,lam)
% Conduct automatic log transformation lam(yy/isig+lam)
% Inputs:
% - y0 [double] series to transform
% - xdir [char] string indating the type of transformation:
% - log: standard log transformation
% - minuslog: log of minus (y0)
% - logsquared: log of y0^2
% - logskew: log of y0 shifted by lam
% - isig [double] scaling factor for y0
% - lam [double] shifting for y0
%
% Outputs:
% - yy [double] transformed series
% - xdir [char] string indating the type of transformation:
% - log: standard log transformation
% - minuslog: log of minus (y0)
% - logsquared: log of y0^2
% - logskew: log of y0 shifted by lam
% - isig [double] scaling factor for y0
% - lam [double] shifting for y0
%
% Notes: takes either one or four arguments. For one argument, the log
% transformation is conducted. For four arguments, the inverse
% transformation is applied.
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
@ -33,12 +58,16 @@ if nargin==1
end
f=@(lam,y)gsa_skewness(log(y+lam));
isig=1;
if ~(max(y0)<0 | min(y0)>0)
if gsa_skewness(y0)<0,
if ~(max(y0)<0 || min(y0)>0)
if gsa_skewness(y0)<0
isig=-1;
y0=-y0;
end
n=hist(y0,10);
if isoctave
n=hist(y0,10);
else
n=histcounts(y0,10);
end
if n(1)>20*n(end)
try
lam=fzero(f,[-min(y0)+10*eps -min(y0)+abs(median(y0))],[],y0);
@ -48,7 +77,7 @@ if ~(max(y0)<0 | min(y0)>0)
if abs(yl(1))<abs(yl(2))
lam=-min(y0)+eps;
else
lam = -min(y0)+abs(median(y0)); %abs(100*(1+min(y0)));
lam = -min(y0)+abs(median(y0));
end
end
yy = log(y0+lam);
@ -63,10 +92,8 @@ else
if max(y0)<0
isig=-1;
y0=-y0;
%yy=log(-y0);
xdir=[xdir0,'_minuslog'];
elseif min(y0)>0
%yy=log(y0);
xdir=[xdir0,'_log'];
end
try
@ -77,7 +104,7 @@ else
if abs(yl(1))<abs(yl(2))
lam=-min(y0)+eps;
else
lam = -min(y0)+abs(median(y0)); %abs(100*(1+min(y0)));
lam = -min(y0)+abs(median(y0));
end
end
lam = max(lam,0);

106
matlab/gsa/map_calibration.m
View File

@ -1,7 +1,13 @@
function map_calibration(OutputDirectoryName, M_, options_, oo_, estim_params_, bayestopt_)
% map_calibration(OutputDirectoryName, M_, options_, oo_, estim_params_, bayestopt_)
% Inputs:
% - OutputDirectoryName [string] name of the output directory
% - M_ [structure] describing the model
% - options_ [structure] describing the options
% - oo_ [structure] storing the results
% - estim_params_ [structure] characterizing parameters to be estimated
% - bayestopt_ [structure] describing the priors
%
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
% marco.ratto@ec.europa.eu
@ -32,16 +38,15 @@ pnames=cell(np,1);
pnames_tex=cell(np,1);
for jj=1:np
if options_.TeX
[param_name_temp, param_name_tex_temp]= get_the_name(nshock+jj, options_.TeX, M_, estim_params_, options_);
pnames_tex{jj,1} = strrep(param_name_tex_temp,'$','');
[param_name_temp, param_name_tex_temp]= get_the_name(nshock+jj, options_.TeX, M_, estim_params_, options_.varobs);
pnames_tex{jj,1} = param_name_tex_temp;
pnames{jj,1} = param_name_temp;
else
param_name_temp = get_the_name(nshock+jj, options_.TeX, M_, estim_params_, options_);
param_name_temp = get_the_name(nshock+jj, options_.TeX, M_, estim_params_, options_.varobs);
pnames{jj,1} = param_name_temp;
end
end
pvalue_ks = options_.opt_gsa.pvalue_ks;
indx_irf = [];
indx_moment = [];
init = ~options_.opt_gsa.load_stab;
@ -63,7 +68,7 @@ if options_.opt_gsa.ppost
filetoload=dir([M_.dname filesep 'metropolis' filesep fname_ '_param_irf*.mat']);
lpmat=[];
for j=1:length(filetoload)
load([M_.dname filesep 'metropolis' filesep fname_ '_param_irf',int2str(j),'.mat'])
load([M_.dname filesep 'metropolis' filesep fname_ '_param_irf',int2str(j),'.mat'],'stock')
lpmat = [lpmat; stock];
clear stock
end
@ -71,12 +76,12 @@ if options_.opt_gsa.ppost
else
if options_.opt_gsa.pprior
filetoload=[OutputDirectoryName '/' fname_ '_prior'];
load(filetoload,'lpmat','lpmat0','istable','iunstable','iindeterm','iwrong' ,'infox')
load(filetoload,'lpmat','lpmat0')
lpmat = [lpmat0 lpmat];
type = 'prior';
else
filetoload=[OutputDirectoryName '/' fname_ '_mc'];
load(filetoload,'lpmat','lpmat0','istable','iunstable','iindeterm','iwrong' ,'infox')
load(filetoload,'lpmat','lpmat0')
lpmat = [lpmat0 lpmat];
type = 'mc';
end
@ -99,17 +104,17 @@ if init
mat_moment{ij}=NaN(Nsam,length(options_.endogenous_prior_restrictions.moment{ij,3}));
end
irestrictions = [1:Nsam];
irestrictions = 1:Nsam;
h = dyn_waitbar(0,'Please wait...');
for j=1:Nsam
M_ = set_all_parameters(lpmat(j,:)',estim_params_,M_);
if nbr_moment_restrictions
[Tt,Rr,SteadyState,info,oo_.dr, M_.params] = dynare_resolve(M_,options_,oo_.dr, oo_.steady_state, oo_.exo_steady_state, oo_.exo_det_steady_state);
[Tt,Rr,~,info,oo_.dr, M_.params] = dynare_resolve(M_,options_,oo_.dr, oo_.steady_state, oo_.exo_steady_state, oo_.exo_det_steady_state);
else
[Tt,Rr,SteadyState,info,oo_.dr, M_.params] = dynare_resolve(M_,options_,oo_.dr, oo_.steady_state, oo_.exo_steady_state, oo_.exo_det_steady_state,'restrict');
[Tt,Rr,~,info,oo_.dr, M_.params] = dynare_resolve(M_,options_,oo_.dr, oo_.steady_state, oo_.exo_steady_state, oo_.exo_det_steady_state,'restrict');
end
if info(1)==0
[info, info_irf, info_moment, data_irf, data_moment]=endogenous_prior_restrictions(Tt,Rr,M_,options_,oo_.dr,oo_.steady_state,oo_.exo_steady_state,oo_.exo_det_steady_state);
[~, info_irf, info_moment, data_irf, data_moment]=endogenous_prior_restrictions(Tt,Rr,M_,options_,oo_.dr,oo_.steady_state,oo_.exo_steady_state,oo_.exo_det_steady_state);
if ~isempty(info_irf)
for ij=1:nbr_irf_restrictions
mat_irf{ij}(j,:)=data_irf{ij}(:,2)';
@ -125,7 +130,9 @@ if init
else
irestrictions(j)=0;
end
dyn_waitbar(j/Nsam,h,['MC iteration ',int2str(j),'/',int2str(Nsam)])
if mod(j,3)==0
dyn_waitbar(j/Nsam,h,['MC iteration ',int2str(j),'/',int2str(Nsam)])
end
end
dyn_waitbar_close(h);
@ -183,7 +190,7 @@ if ~isempty(indx_irf)
maxijv=0;
for ij=1:nbr_irf_restrictions
if length(endo_prior_restrictions.irf{ij,3})>maxijv
maxij=ij;maxijv=length(endo_prior_restrictions.irf{ij,3});
maxijv=length(endo_prior_restrictions.irf{ij,3});
end
plot_indx(ij) = find(strcmp(irf_couples,all_irf_couples(ij,:)));
time_matrix{plot_indx(ij)} = [time_matrix{plot_indx(ij)} endo_prior_restrictions.irf{ij,3}];
@ -235,38 +242,25 @@ if ~isempty(indx_irf)
hold off,
% hold off,
title([endo_prior_restrictions.irf{ij,1},' vs ',endo_prior_restrictions.irf{ij,2}, '(', leg,')'],'interpreter','none'),
%set(legend_h,'Xlim',[0 1]);
% if ij==maxij
% leg1 = num2str(endo_prior_restrictions.irf{ij,3}(:));
% [legend_h,object_h,plot_h,text_strings]=legend(leg1);
% Position=get(legend_h,'Position');Position(1:2)=[-0.055 0.95-Position(4)];
% set(legend_h,'Position',Position);
% end
end
% hc = get(h,'Children');
%for i=2:2:length(hc)
%end
indx1 = find(indx_irf(:,ij)==0);
indx2 = find(indx_irf(:,ij)~=0);
atitle0=[endo_prior_restrictions.irf{ij,1},' vs ',endo_prior_restrictions.irf{ij,2}, '(', leg,')'];
fprintf(['%4.1f%% of the ',type,' support matches IRF ',atitle0,' inside [%4.1f, %4.1f]\n'],length(indx1)/length(irestrictions)*100,endo_prior_restrictions.irf{ij,4})
% aname=[type '_irf_calib_',int2str(ij)];
aname=[type '_irf_calib_',endo_prior_restrictions.irf{ij,1},'_vs_',endo_prior_restrictions.irf{ij,2},'_',aleg];
atitle=[type ' IRF Calib: Parameter(s) driving ',endo_prior_restrictions.irf{ij,1},' vs ',endo_prior_restrictions.irf{ij,2}, '(', leg,')'];
options_mcf.amcf_name = aname;
options_mcf.amcf_title = atitle;
options_mcf.beha_title = 'IRF restriction';
options_mcf.nobeha_title = 'NO IRF restriction';
if options_.TeX
options_mcf.beha_title_latex = 'IRF restriction';
options_mcf.nobeha_title_latex = 'NO IRF restriction';
end
options_mcf.title = atitle0;
if ~isempty(indx1) && ~isempty(indx2)
mcf_analysis(xmat(:,nshock+1:end), indx1, indx2, options_mcf, options_);
mcf_analysis(xmat(:,nshock+1:end), indx1, indx2, options_mcf, M_, options_, bayestopt_, estim_params_);
end
% [proba, dproba] = stab_map_1(xmat, indx1, indx2, aname, 0);
% indplot=find(proba<pvalue_ks);
% if ~isempty(indplot)
% stab_map_1(xmat, indx1, indx2, aname, 1, indplot, OutputDirectoryName,[],atitle);
% end
end
for ij=1:nbr_irf_couples
if length(time_matrix{ij})>1
@ -284,7 +278,6 @@ if ~isempty(indx_irf)
tmp(temp_index,:) = endo_prior_restrictions.irf{itmp(ir),4};
end
end
% tmp = cell2mat(endo_prior_restrictions.irf(itmp,4));
tmp(isinf(tmp(:,1)),1)=a(3);
tmp(isinf(tmp(:,2)),2)=a(4);
hp = patch([time_matrix{ij} time_matrix{ij}(end:-1:1)],[tmp(:,1); tmp(end:-1:1,2)],'c');
@ -297,7 +290,6 @@ if ~isempty(indx_irf)
hold off
axis([max(1,a(1)) a(2:4)])
box on
%set(gca,'xtick',sort(time_matrix{ij}))
itmp = min(itmp);
title([endo_prior_restrictions.irf{itmp,1},' vs ',endo_prior_restrictions.irf{itmp,2}],'interpreter','none'),
end
@ -311,16 +303,20 @@ if ~isempty(indx_irf)
aleg = 'ALL';
atitle0=[endo_prior_restrictions.irf{itmp,1},' vs ',endo_prior_restrictions.irf{itmp,2}, '(', leg,')'];
fprintf(['%4.1f%% of the ',type,' support matches IRF restrictions ',atitle0,'\n'],length(indx1)/length(irestrictions)*100)
% aname=[type '_irf_calib_',int2str(ij)];
aname=[type '_irf_calib_',endo_prior_restrictions.irf{itmp,1},'_vs_',endo_prior_restrictions.irf{itmp,2},'_',aleg];
atitle=[type ' IRF Calib: Parameter(s) driving ',endo_prior_restrictions.irf{itmp,1},' vs ',endo_prior_restrictions.irf{itmp,2}, '(', leg,')'];
options_mcf.amcf_name = aname;
options_mcf.amcf_title = atitle;
options_mcf.beha_title = 'IRF restriction';
options_mcf.nobeha_title = 'NO IRF restriction';
if options_.TeX
options_mcf.beha_title_latex = 'IRF restriction';
options_mcf.nobeha_title_latex = 'NO IRF restriction';
end
options_mcf.title = atitle0;
if ~isempty(indx1) && ~isempty(indx2)
mcf_analysis(xmat(:,nshock+1:end), indx1, indx2, options_mcf, options_);
mcf_analysis(xmat(:,nshock+1:end), indx1, indx2, options_mcf, M_, options_, bayestopt_, estim_params_);
end
end
end
@ -368,11 +364,11 @@ if ~isempty(indx_moment)
name_tex=cell(np,1);
for jj=1:np
if options_.TeX
[param_name_temp, param_name_tex_temp]= get_the_name(jj,options_.TeX,M_,estim_params_,options_);
name_tex{jj,1} = strrep(param_name_tex_temp,'$','');
[param_name_temp, param_name_tex_temp]= get_the_name(jj,options_.TeX,M_,estim_params_,options_.varobs);
name_tex{jj,1} = param_name_tex_temp;
name{jj,1} = param_name_temp;
else
param_name_temp = get_the_name(jj,options_.TeX,M_,estim_params_,options_);
param_name_temp = get_the_name(jj,options_.TeX,M_,estim_params_,options_.varobs);
name{jj,1} = param_name_temp;
end
end
@ -398,7 +394,7 @@ if ~isempty(indx_moment)
for ij=1:nbr_moment_restrictions
endo_prior_restrictions.moment{ij,3} = sort(endo_prior_restrictions.moment{ij,3});
if length(endo_prior_restrictions.moment{ij,3})>maxijv
maxij=ij;maxijv=length(endo_prior_restrictions.moment{ij,3});
maxijv=length(endo_prior_restrictions.moment{ij,3});
end
plot_indx(ij) = find(strcmp(moment_couples,all_moment_couples(ij,:)));
time_matrix{plot_indx(ij)} = [time_matrix{plot_indx(ij)} endo_prior_restrictions.moment{ij,3}];
@ -450,34 +446,25 @@ if ~isempty(indx_moment)
set(hc,'color','k','linewidth',2)
hold off
title([endo_prior_restrictions.moment{ij,1},' vs ',endo_prior_restrictions.moment{ij,2},'(',leg,')'],'interpreter','none'),
% if ij==maxij
% leg1 = num2str(endo_prior_restrictions.moment{ij,3}(:));
% [legend_h,object_h,plot_h,text_strings]=legend(leg1);
% Position=get(legend_h,'Position');Position(1:2)=[-0.055 0.95-Position(4)];
% set(legend_h,'Position',Position);
% end
end
indx1 = find(indx_moment(:,ij)==0);
indx2 = find(indx_moment(:,ij)~=0);
atitle0=[endo_prior_restrictions.moment{ij,1},' vs ',endo_prior_restrictions.moment{ij,2}, '(', leg,')'];
fprintf(['%4.1f%% of the ',type,' support matches MOMENT ',atitle0,' inside [%4.1f, %4.1f]\n'],length(indx1)/length(irestrictions)*100,endo_prior_restrictions.moment{ij,4})
% aname=[type '_moment_calib_',int2str(ij)];
aname=[type '_moment_calib_',endo_prior_restrictions.moment{ij,1},'_vs_',endo_prior_restrictions.moment{ij,2},'_',aleg];
atitle=[type ' MOMENT Calib: Parameter(s) driving ',endo_prior_restrictions.moment{ij,1},' vs ',endo_prior_restrictions.moment{ij,2}, '(', leg,')'];
options_mcf.amcf_name = aname;
options_mcf.amcf_title = atitle;
options_mcf.beha_title = 'moment restriction';
options_mcf.nobeha_title = 'NO moment restriction';
if options_.TeX
options_mcf.beha_title_latex = 'moment restriction';
options_mcf.nobeha_title_latex = 'NO moment restriction';
end
options_mcf.title = atitle0;
if ~isempty(indx1) && ~isempty(indx2)
mcf_analysis(xmat, indx1, indx2, options_mcf, options_);
mcf_analysis(xmat, indx1, indx2, options_mcf, M_, options_, bayestopt_, estim_params_);
end
% [proba, dproba] = stab_map_1(xmat, indx1, indx2, aname, 0);
% indplot=find(proba<pvalue_ks);
% if ~isempty(indplot)
% stab_map_1(xmat, indx1, indx2, aname, 1, indplot, OutputDirectoryName,[],atitle);
% end
end
for ij=1:nbr_moment_couples
time_matrix{ij} = sort(time_matrix{ij});
@ -496,7 +483,6 @@ if ~isempty(indx_moment)
tmp(temp_index,:) = endo_prior_restrictions.moment{itmp(ir),4};
end
end
% tmp = cell2mat(endo_prior_restrictions.moment(itmp,4));
tmp(isinf(tmp(:,1)),1)=a(3);
tmp(isinf(tmp(:,2)),2)=a(4);
hp = patch([time_matrix{ij} time_matrix{ij}(end:-1:1)],[tmp(:,1); tmp(end:-1:1,2)],'b');
@ -509,7 +495,6 @@ if ~isempty(indx_moment)
hold off
axis(a)
box on
% set(gca,'xtick',sort(time_matrix{ij}))
itmp = min(itmp);
title([endo_prior_restrictions.moment{itmp,1},' vs ',endo_prior_restrictions.moment{itmp,2}],'interpreter','none'),
end
@ -523,16 +508,19 @@ if ~isempty(indx_moment)
aleg = 'ALL';
atitle0=[endo_prior_restrictions.moment{itmp,1},' vs ',endo_prior_restrictions.moment{itmp,2}, '(', leg,')'];
fprintf(['%4.1f%% of the ',type,' support matches MOMENT restrictions ',atitle0,'\n'],length(indx1)/length(irestrictions)*100)
% aname=[type '_moment_calib_',int2str(ij)];
aname=[type '_moment_calib_',endo_prior_restrictions.moment{itmp,1},'_vs_',endo_prior_restrictions.moment{itmp,2},'_',aleg];
atitle=[type ' MOMENT Calib: Parameter(s) driving ',endo_prior_restrictions.moment{itmp,1},' vs ',endo_prior_restrictions.moment{itmp,2}, '(', leg,')'];
options_mcf.amcf_name = aname;
options_mcf.amcf_title = atitle;
options_mcf.beha_title = 'moment restriction';
options_mcf.nobeha_title = 'NO moment restriction';
if options_.TeX
options_mcf.beha_title_latex = 'moment restriction';
options_mcf.nobeha_title_latex = 'NO moment restriction';
end
options_mcf.title = atitle0;
if ~isempty(indx1) && ~isempty(indx2)
mcf_analysis(xmat, indx1, indx2, options_mcf, options_);
mcf_analysis(xmat, indx1, indx2, options_mcf, M_, options_, bayestopt_, estim_params_);
end
end
end

1383
matlab/gsa/map_ident_.m
View File

File diff suppressed because it is too large Load Diff

39
matlab/gsa/mc_moments.m
View File

@ -1,6 +1,19 @@
function [vdec, cc, ac] = mc_moments(mm, ss, dr)
function [vdec, cc, ac] = mc_moments(mm, ss, dr, M_, options_, estim_params_)
% [vdec, cc, ac] = mc_moments(mm, ss, dr, M_, options_,estim_params_)
% Conduct Monte Carlo simulation of second moments for GSA
% Inputs:
% - dr [structure] decision rules
% - M_ [structure] model structure
% - options_ [structure] Matlab's structure describing the current options
% - estim_params_ [structure] characterizing parameters to be estimated
%
% Outputs:
% - vdec [double] variance decomposition matrix
% - cc [double] vector of unique elements of cross correlation matrix
% - ac [cell] autocorrelation matrix
% Copyright © 2012-2018 Dynare Team
% Copyright © 2012-2023 Dynare Team
%
% This file is part of Dynare.
%
@ -17,30 +30,30 @@ function [vdec, cc, ac] = mc_moments(mm, ss, dr)
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
global options_ M_ estim_params_ oo_
[nr1, nc1, nsam] = size(mm);
[~, nc1, nsam] = size(mm);
nobs=length(options_.varobs);
disp('Computing theoretical moments ...')
disp('mc_moments: Computing theoretical moments ...')
h = dyn_waitbar(0,'Theoretical moments ...');
vdec = zeros(nobs,M_.exo_nbr,nsam);
cc = zeros(nobs,nobs,nsam);
ac = zeros(nobs,nobs*options_.ar,nsam);
for j=1:nsam
oo_.dr.ghx = mm(:, [1:(nc1-M_.exo_nbr)],j);
oo_.dr.ghu = mm(:, [(nc1-M_.exo_nbr+1):end], j);
dr.ghx = mm(:, 1:(nc1-M_.exo_nbr),j);
dr.ghu = mm(:, (nc1-M_.exo_nbr+1):end, j);
if ~isempty(ss)
set_shocks_param(ss(j,:));
M_=set_shocks_param(M_,estim_params_,ss(j,:));
end
[vdec(:,:,j), corr, autocorr, z, zz] = th_moments(oo_.dr,options_.varobs);
[vdec(:,:,j), corr, autocorr] = th_moments(dr,options_,M_);
cc(:,:,j)=triu(corr);
dum=[];
dum=NaN(nobs,nobs*options_.ar);
for i=1:options_.ar
dum=[dum, autocorr{i}];
dum(:,(i-1)*nobs+1:i*nobs)=autocorr{i};
end
ac(:,:,j)=dum;
dyn_waitbar(j/nsam,h)
if mod(j,3)==0
dyn_waitbar(j/nsam,h)
end
end
dyn_waitbar_close(h)
skipline()

52
matlab/gsa/mcf_analysis.m
View File

@ -1,5 +1,17 @@
function indmcf = mcf_analysis(lpmat, ibeha, inobeha, options_mcf, options_)
% indmcf = mcf_analysis(lpmat, ibeha, inobeha, options_mcf, options_)
function indmcf = mcf_analysis(lpmat, ibeha, inobeha, options_mcf, M_, options_, bayestopt_, estim_params_)
% indmcf = mcf_analysis(lpmat, ibeha, inobeha, options_mcf, M_, options_, bayestopt_, estim_params_)
% Inputs:
% - lpmat [double] Monte Carlo matrix
% - ibeha [integer] index of behavioural runs
% - inobeha [integer] index of non-behavioural runs
% - options_gsa_ [structure] GSA options_
% - M_ [structure] describing the model
% - options_ [structure] describing the options
% - bayestopt_ [structure] describing the priors
% - estim_params_ [structure] characterizing parameters to be estimated
%
% Outputs:
% - indmcf [double] results of matrix
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
@ -42,6 +54,10 @@ amcf_name = options_mcf.amcf_name;
amcf_title = options_mcf.amcf_title;
beha_title = options_mcf.beha_title;
nobeha_title = options_mcf.nobeha_title;
if options_.TeX
beha_title_latex = options_mcf.beha_title_latex;
nobeha_title_latex = options_mcf.nobeha_title_latex;
end
title = options_mcf.title;
fname_ = options_mcf.fname_;
xparam1=[];
@ -50,15 +66,15 @@ if isfield(options_mcf,'xparam1')
end
OutputDirectoryName = options_mcf.OutputDirectoryName;
[proba, dproba] = stab_map_1(lpmat, ibeha, inobeha, [],0);
[proba, dproba] = stab_map_1(lpmat, ibeha, inobeha, [],fname_, options_, bayestopt_.name, estim_params_,0);
indmcf=find(proba<pvalue_ks);
[tmp,jtmp] = sort(proba(indmcf),2,'ascend');
[~,jtmp] = sort(proba(indmcf),1,'ascend');
indmcf = indmcf(jtmp);
if ~isempty(indmcf)
skipline()
headers = {'Parameter','d-stat','p-value'};
labels = param_names(indmcf);
data_mat=[dproba(indmcf)' proba(indmcf)'];
data_mat=[dproba(indmcf) proba(indmcf)];
options_temp.noprint=0;
dyntable(options_temp,['Smirnov statistics in driving ', title],headers,labels,data_mat,size(labels,2)+2,16,3);
if options_.TeX
@ -69,10 +85,14 @@ if ~isempty(indmcf)
end
end
if length(ibeha)>10 && length(inobeha)>10
indcorr1 = stab_map_2(lpmat(ibeha,:),alpha2, pvalue_corr, beha_title);
indcorr2 = stab_map_2(lpmat(inobeha,:),alpha2, pvalue_corr, nobeha_title);
if options_.TeX
indcorr1 = stab_map_2(lpmat(ibeha,:),alpha2, pvalue_corr, M_, options_, bayestopt_, estim_params_, beha_title, beha_title_latex);
indcorr2 = stab_map_2(lpmat(inobeha,:),alpha2, pvalue_corr, M_, options_, bayestopt_, estim_params_, nobeha_title, nobeha_title_latex);
else
indcorr1 = stab_map_2(lpmat(ibeha,:),alpha2, pvalue_corr, M_, options_, bayestopt_, estim_params_, beha_title);
indcorr2 = stab_map_2(lpmat(inobeha,:),alpha2, pvalue_corr, M_, options_, bayestopt_, estim_params_, nobeha_title);
end
indcorr = union(indcorr1(:), indcorr2(:));
indcorr = indcorr(~ismember(indcorr(:),indmcf));
indmcf = [indmcf(:); indcorr(:)];
@ -80,8 +100,16 @@ end
if ~isempty(indmcf) && ~options_.nograph
skipline()
xx=[];
if ~ isempty(xparam1), xx=xparam1(indmcf); end
scatter_mcf(lpmat(ibeha,indmcf),lpmat(inobeha,indmcf), param_names_tex(indmcf), ...
'.', [fname_,'_',amcf_name], OutputDirectoryName, amcf_title,xx, options_, ...
beha_title, nobeha_title)
if ~ isempty(xparam1)
xx=xparam1(indmcf);
end
if options_.TeX
scatter_mcf(lpmat(ibeha,indmcf),lpmat(inobeha,indmcf), param_names_tex(indmcf), ...
'.', [fname_,'_',amcf_name], OutputDirectoryName, amcf_title,xx, options_, ...
beha_title, nobeha_title, beha_title_latex, nobeha_title_latex)
else
scatter_mcf(lpmat(ibeha,indmcf),lpmat(inobeha,indmcf), param_names_tex(indmcf), ...
'.', [fname_,'_',amcf_name], OutputDirectoryName, amcf_title,xx, options_, ...
beha_title, nobeha_title)
end
end

113
matlab/gsa/myboxplot.m
View File

@ -1,12 +1,8 @@
function sout = myboxplot (data,notched,symbol,vertical,maxwhisker)
% sout = myboxplot (data,notched,symbol,vertical,maxwhisker)
% sout = myboxplot (data,notched,symbol,vertical,maxwhisker)
% Creates a box plot
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
% marco.ratto@ec.europa.eu
% Copyright © 2012 European Commission
% Copyright © 2012-2017 Dynare Team
% Copyright © 2010-2023 Dynare Team
%
% This file is part of Dynare.
%
@ -23,18 +19,17 @@ function sout = myboxplot (data,notched,symbol,vertical,maxwhisker)
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
% % % % endif
if nargin < 5 | isempty(maxwhisker), maxwhisker = 1.5; end
if nargin < 4 | isempty(vertical), vertical = 1; end
if nargin < 3 | isempty(symbol), symbol = ['+','o']; end
if nargin < 2 | isempty(notched), notched = 0; end
if nargin < 5 || isempty(maxwhisker), maxwhisker = 1.5; end
if nargin < 4 || isempty(vertical), vertical = 1; end
if nargin < 3 || isempty(symbol), symbol = ['+','o']; end
if nargin < 2 || isempty(notched), notched = 0; end
if length(symbol)==1, symbol(2)=symbol(1); end
if notched==1, notched=0.25; end
a=1-notched;
% ## figure out how many data sets we have
% % figure out how many data sets we have
if iscell(data)
nc = length(data);
else
@ -42,11 +37,11 @@ else
nc = size(data,2);
end
% ## compute statistics
% ## s will contain
% ## 1,5 min and max
% ## 2,3,4 1st, 2nd and 3rd quartile
% ## 6,7 lower and upper confidence intervals for median
% compute statistics
% s will contain
% 1,5 min and max
% 2,3,4 1st, 2nd and 3rd quartile
% 6,7 lower and upper confidence intervals for median
s = zeros(7,nc);
box = zeros(1,nc);
whisker_x = ones(2,1)*[1:nc,1:nc];
@ -57,44 +52,36 @@ outliers2_x = [];
outliers2_y = [];
for i=1:nc
% ## Get the next data set from the array or cell array
% Get the next data set from the array or cell array
if iscell(data)
col = data{i}(:);
else
col = data(:,i);
end
% ## Skip missing data
% Skip missing data
% % % % % % % col(isnan(col) | isna (col)) = [];
col(isnan(col)) = [];
% ## Remember the data length
% Remember the data length
nd = length(col);
box(i) = nd;
if (nd > 1)
% ## min,max and quartiles
% s(1:5,i) = statistics(col)(1:5);
% min,max and quartiles
s(1,i)=min(col);
s(5,i)=max(col);
s(2,i)=myprctilecol(col,25);
s(3,i)=myprctilecol(col,50);
s(4,i)=myprctilecol(col,75);
% ## confidence interval for the median
% confidence interval for the median
est = 1.57*(s(4,i)-s(2,i))/sqrt(nd);
s(6,i) = max([s(3,i)-est, s(2,i)]);
s(7,i) = min([s(3,i)+est, s(4,i)]);
% ## whiskers out to the last point within the desired inter-quartile range
% whiskers out to the last point within the desired inter-quartile range
IQR = maxwhisker*(s(4,i)-s(2,i));
whisker_y(:,i) = [min(col(col >= s(2,i)-IQR)); s(2,i)];
whisker_y(:,nc+i) = [max(col(col <= s(4,i)+IQR)); s(4,i)];
% ## outliers beyond 1 and 2 inter-quartile ranges
% outliers beyond 1 and 2 inter-quartile ranges
outliers = col((col < s(2,i)-IQR & col >= s(2,i)-2*IQR) | (col > s(4,i)+IQR & col <= s(4,i)+2*IQR));
outliers2 = col(col < s(2,i)-2*IQR | col > s(4,i)+2*IQR);
outliers_x = [outliers_x; i*ones(size(outliers))];
@ -102,41 +89,37 @@ for i=1:nc
outliers2_x = [outliers2_x; i*ones(size(outliers2))];
outliers2_y = [outliers2_y; outliers2];
elseif (nd == 1)
% ## all statistics collapse to the value of the point
% all statistics collapse to the value of the point
s(:,i) = col;
% ## single point data sets are plotted as outliers.
% single point data sets are plotted as outliers.
outliers_x = [outliers_x; i];
outliers_y = [outliers_y; col];
else
% ## no statistics if no points
% no statistics if no points
s(:,i) = NaN;
end
end
% % % % if isempty(outliers2_y)
% % % % outliers2_y=
% ## Note which boxes don't have enough stats
% Note which boxes don't have enough stats
chop = find(box <= 1);
% ## Draw a box around the quartiles, with width proportional to the number of
% ## items in the box. Draw notches if desired.
% Draw a box around the quartiles, with width proportional to the number of
% items in the box. Draw notches if desired.
box = box*0.23/max(box);
quartile_x = ones(11,1)*[1:nc] + [-a;-1;-1;1;1;a;1;1;-1;-1;-a]*box;
quartile_y = s([3,7,4,4,7,3,6,2,2,6,3],:);
% ## Draw a line through the median
% Draw a line through the median
median_x = ones(2,1)*[1:nc] + [-a;+a]*box;
% median_x=median(col);
median_y = s([3,3],:);
% ## Chop all boxes which don't have enough stats
% Chop all boxes which don't have enough stats
quartile_x(:,chop) = [];
quartile_y(:,chop) = [];
whisker_x(:,[chop,chop+nc]) = [];
whisker_y(:,[chop,chop+nc]) = [];
median_x(:,chop) = [];
median_y(:,chop) = [];
% % % %
% ## Add caps to the remaining whiskers
% Add caps to the remaining whiskers
cap_x = whisker_x;
cap_x(1,:) =cap_x(1,:)- 0.05;
cap_x(2,:) =cap_x(2,:)+ 0.05;
@ -146,11 +129,14 @@ cap_y = whisker_y([1,1],:);
% #whisker_x,whisker_y
% #median_x,median_y
% #cap_x,cap_y
%
% ## Do the plot
% Do the plot
mm=min(min(data));
MM=max(max(data));
if isnan(mm)
mm=0;
MM=0;
end
if vertical
plot (quartile_x, quartile_y, 'b', ...
@ -162,17 +148,30 @@ if vertical
set(gca,'XTick',1:nc);
set(gca, 'XLim', [0.5, nc+0.5]);
set(gca, 'YLim', [mm-(MM-mm)*0.05-eps, MM+(MM-mm)*0.05+eps]);
else
% % % % % plot (quartile_y, quartile_x, "b;;",
% % % % % whisker_y, whisker_x, "b;;",
% % % % % cap_y, cap_x, "b;;",
% % % % % median_y, median_x, "r;;",
% % % % % outliers_y, outliers_x, [symbol(1),"r;;"],
% % % % % outliers2_y, outliers2_x, [symbol(2),"r;;"]);
end
if nargout
sout=s;
end
% % % endfunction
function y = myprctilecol(x,p)
xx = sort(x);
[m,n] = size(x);
if m==1 | n==1
m = max(m,n);
if m == 1
y = x*ones(length(p),1);
return
end
n = 1;
q = 100*(0.5:m - 0.5)./m;
xx = [min(x); xx(:); max(x)];
else
q = 100*(0.5:m - 0.5)./m;
xx = [min(x); xx; max(x)];
end
q = [0 q 100];
y = interp1(q,xx,p);

43
matlab/gsa/myprctilecol.m
View File

@ -1,43 +0,0 @@
function y = myprctilecol(x,p)
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
% marco.ratto@ec.europa.eu
% Copyright © 2012 European Commission
% Copyright © 2012-2017 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 <https://www.gnu.org/licenses/>.
xx = sort(x);
[m,n] = size(x);
if m==1 | n==1
m = max(m,n);
if m == 1
y = x*ones(length(p),1);
return
end
n = 1;
q = 100*(0.5:m - 0.5)./m;
xx = [min(x); xx(:); max(x)];
else
q = 100*(0.5:m - 0.5)./m;
xx = [min(x); xx; max(x)];
end
q = [0 q 100];
y = interp1(q,xx,p);

91
matlab/gsa/pick.m
View File

@ -1,91 +0,0 @@
function pick
%
% Copyright © 2001-2017 European Commission
% Copyright © 2017-2023 DynareTeam
% This file is part of GLUEWIN
% GLUEWIN is a MATLAB code designed for analysing the output
% of Monte Carlo runs when empirical observations of the model output are available
% and implements the GSA-GLUE methodology by Ratto et al. [1], based on a combination
% of GLUE (Generalised Likelihood Uncertainty Estimation) by K. Beven [2] and GSA
% Global Sensitivity Analysis) [3].']
% The program has been developed by M. Ratto, European Commission, Joint Research Centre,
% Institute for the Protection and Security of The Citizen, Technological and Economic Risk Management,
% Applied Statistics, as a deliverable of the IMPACT project
% (EC Fifth Framework Programme, SCA Project, IST-1999-11313, DG-INFSO).
%
% The graphical layout of the code is inspired by the freeware GLUE package by K. Beven,
% vailable at the Lancaster University web site on the page [4]:
% http://www.es.lancs.ac.uk/hfdg/glue.html
% to which the GLUEWIN code introduces several extensions and additional options.
% Thanks are due to R. Girardi, A. Rossi, A. Saltelli, S. Tarantola and U. Callies for numerous
% comments and suggestions.
% For more information, please contact marco.ratto@ec.europa.eu
%
% Disclaimer: This software has been developed at the Joint Research Centre of European Commission
% by officers in the course of their official duties. This software is not subject to copyright
% protection and is in the public domain. It is an experimental system. The Joint Research Centre
% of European Commission assumes no responsibility whatsoever for its use by other parties
% and makes no guarantees, expressed or implied, about its quality, reliability, or any other
% characteristic. We would appreciate acknowledgement if the software is used.
%
% [1] Ratto, M., Tarantola, S., A. Saltelli, Sensitivity analysis in model calibration: GSA-GLUE approach.
% 'Computer Physics Communications, 136, 2001, 212-224
% [2] Beven K.J., Binley A., The Future of Distributed Models: Model Calibration and Uncertainty
% 'Prediction, Hydrological Processes, 6, 279-298, 1992
% [3] Saltelli, A., K. Chan, M. Scott, Editors, (2000), Sensitivity analysis, John Wiley & Sons
% 'publishers, Probability and Statistics series.
% [4] Beven K., GLUE for Windows User manual, 1998.
pmenu=findobj(gcf,'type','uicontextmenu','Tag','Run viewer');
button1=findobj(gcf,'type','uimenu','Tag','save params');
button2=findobj(gcf,'type','uimenu','Tag','eval params');
%button=get(pmenu,'children');
gg=gco;
ax0=gca;
set(gg,'buttondownfcn',[]);
c=get(gca,'currentpoint');
x=c(1,1);
y=c(1,2);
X=get(gco,'xdata');
Y=get(gco,'ydata');
dx=get(gca,'xlim');
dy=get(gca,'ylim');
pos=get(gca,'position');
scalex=dx(2)-dx(1);
scaley=dy(2)-dy(1);
if length(X)>1
K = dsearchn([(Y./scaley)' (X./scalex)'],[y/scaley x/scalex]);
else
az=get(gca,'children');
T =get(az(end),'ydata');
[dum K]=max(T);
end
KK=K;
set(button1,'Label',['Save ',num2str(K)],'Callback',['scatter_callback(',num2str(KK),',''save'')']);
set(button2,'Label',['Eval ',num2str(K)],'Callback',['scatter_callback(',num2str(KK),',''eval'')']);
hh_obj=findobj(gcf,'type','axes','Tag','scatter');
for k=1:length(hh_obj)
axes(hh_obj(k));
dum=get(gca,'children');
dumx=get(dum(end),'xdata');
dumy=get(dum(end),'ydata');
xmid=min(dumx) + 0.5*(max(dumx)-min(dumx));
hold on
plot(dumx(KK),dumy(KK),'or');
if dumx(KK) < xmid
text(dumx(KK),dumy(KK),[' ',num2str(K)], ...
'FontWeight','Bold',...
'Color','r');
else
text(dumx(KK),dumy(KK),[num2str(K),' '], ...
'HorizontalAlignment','right', ...
'FontWeight','Bold',...
'Color','r');
end
hold off
end

39
matlab/gsa/prior_draw_gsa.m
View File

@ -1,9 +1,13 @@
function pdraw = prior_draw_gsa(init,rdraw)
function pdraw = prior_draw_gsa(M_,bayestopt_,options_,estim_params_,init,rdraw)
% Draws from the prior distributions for use with Sensitivity Toolbox for DYNARE
%
% INPUTS
% o init [integer] scalar equal to 1 (first call) or 0.
% o rdraw
% - M_ [structure] describing the model
% - bayestopt_ [structure] describing the priors
% - options_ [structure] describing the options
% - estim_params_ [structure] characterizing parameters to be estimated
% - init [integer] scalar equal to 1 (first call) or 0.
% - rdraw
%
% OUTPUTS
% o pdraw [double] draw from the joint prior density.
@ -35,8 +39,7 @@ function pdraw = prior_draw_gsa(init,rdraw)
%
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
global bayestopt_ options_ estim_params_ M_
persistent npar pshape p6 p7 p3 p4 lbcum ubcum
if init
@ -49,7 +52,7 @@ if init
pdraw = zeros(npar,1);
lbcum = zeros(npar,1);
ubcum = ones(npar,1);
[~,~,~,lb,ub,~] = set_prior(estim_params_,M_,options_); %Prepare bounds
[~,~,~,lb,ub] = set_prior(estim_params_,M_,options_); %Prepare bounds
if ~isempty(bayestopt_) && any(bayestopt_.pshape > 0)
% Set prior bounds
bounds = prior_bounds(bayestopt_, options_.prior_trunc);
@ -64,29 +67,29 @@ if init
% set bounds for cumulative probabilities
for i = 1:npar
switch pshape(i)
case 5% Uniform prior.
p4(i) = min(p4(i),bounds.ub(i));
p3(i) = max(p3(i),bounds.lb(i));
case 3% Gaussian prior.
lbcum(i) = 0.5 * erfc(-(bounds.lb(i)-p6(i))/p7(i) ./ sqrt(2));
ubcum(i) = 0.5 * erfc(-(bounds.ub(i)-p6(i))/p7(i) ./ sqrt(2));
case 2% Gamma prior.
lbcum(i) = gamcdf(bounds.lb(i)-p3(i),p6(i),p7(i));
ubcum(i) = gamcdf(bounds.ub(i)-p3(i),p6(i),p7(i));
case 1% Beta distribution (TODO: generalized beta distribution)
lbcum(i) = betainc((bounds.lb(i)-p3(i))./(p4(i)-p3(i)),p6(i),p7(i));
ubcum(i) = betainc((bounds.ub(i)-p3(i))./(p4(i)-p3(i)),p6(i),p7(i));
case 2% Gamma prior.
lbcum(i) = gamcdf(bounds.lb(i)-p3(i),p6(i),p7(i));
ubcum(i) = gamcdf(bounds.ub(i)-p3(i),p6(i),p7(i));
case 3% Gaussian prior.
lbcum(i) = 0.5 * erfc(-(bounds.lb(i)-p6(i))/p7(i) ./ sqrt(2));
ubcum(i) = 0.5 * erfc(-(bounds.ub(i)-p6(i))/p7(i) ./ sqrt(2));
case 4% INV-GAMMA1 distribution
% TO BE CHECKED
lbcum(i) = gamcdf(1/(bounds.ub(i)-p3(i))^2,p7(i)/2,2/p6(i));
ubcum(i) = gamcdf(1/(bounds.lb(i)-p3(i))^2,p7(i)/2,2/p6(i));
case 5% Uniform prior.
p4(i) = min(p4(i),bounds.ub(i));
p3(i) = max(p3(i),bounds.lb(i));
case 6% INV-GAMMA2 distribution
% TO BE CHECKED
lbcum(i) = gamcdf(1/(bounds.ub(i)-p3(i)),p7(i)/2,2/p6(i));
ubcum(i) = gamcdf(1/(bounds.lb(i)-p3(i)),p7(i)/2,2/p6(i));
case 8
lbcum(i) = weibcdf(bounds.lb(i)-p3(i),p6(i),p7(i));
ubcum(i) = weibcdf(bounds.ub(i)-p3(i),p6(i),p7(i));
lbcum(i) = wblcdf(bounds.lb(i)-p3(i),p6(i),p7(i));
ubcum(i) = wblcdf(bounds.ub(i)-p3(i),p6(i),p7(i));
otherwise
% Nothing to do here.
end
@ -94,7 +97,7 @@ if init
return
end
pdraw=NaN(size(rdraw,1),npar);
for i = 1:npar
rdraw(:,i) = rdraw(:,i).*(ubcum(i)-lbcum(i))+lbcum(i);
switch pshape(i)

5
matlab/gsa/priorcdf.m
View File

@ -1,5 +1,5 @@
function xcum = priorcdf(para, pshape, p6, p7, p3, p4)
% xcum = priorcdf(para, pshape, p6, p7, p3, p4)
% This procedure transforms x vectors into cumulative values
% pshape: 0 is point mass, both para and p2 are ignored
% 1 is BETA(mean,stdd)
@ -11,7 +11,7 @@ function xcum = priorcdf(para, pshape, p6, p7, p3, p4)
% 8 is WEIBULL(s, k)
% Adapted by M. Ratto from MJ priordens.m
% Copyright © 2012-2015 Dynare Team
% Copyright © 2012-2023 Dynare Team
%
% This file is part of Dynare.
%
@ -28,6 +28,7 @@ function xcum = priorcdf(para, pshape, p6, p7, p3, p4)
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
xcum=NaN(size(para));
for i=1:length(pshape)
switch pshape(i)
case 1 % (generalized) BETA Prior

44
matlab/gsa/read_data.m
View File

@ -1,44 +0,0 @@
function [gend, data] = read_data()
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
% marco.ratto@ec.europa.eu
% Copyright © 2012-2015 European Commission
% Copyright © 2012-2017 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 <https://www.gnu.org/licenses/>.
global options_
rawdata = read_variables(options_.datafile,options_.varobs,[],options_.xls_sheet,options_.xls_range);
options_ = set_default_option(options_,'nobs',size(rawdata,1)-options_.first_obs+1);
gend = options_.nobs;
rawdata = rawdata(options_.first_obs:options_.first_obs+gend-1,:);
if options_.loglinear == 1 & ~options_.logdata
rawdata = log(rawdata);
end
if options_.prefilter == 1
data = transpose(rawdata-ones(gend,1)* mean(rawdata,1));
else
data = transpose(rawdata);
end
if ~isreal(rawdata)
error(['There are complex values in the data. Probably a wrong' ...
' transformation'])
end

280
matlab/gsa/redform_map.m
View File

@ -1,15 +1,13 @@
function redform_map(dirname,options_gsa_)
%function redform_map(dirname)
% inputs (from opt_gsa structure
% anamendo = options_gsa_.namendo;
% anamlagendo = options_gsa_.namlagendo;
% anamexo = options_gsa_.namexo;
% iload = options_gsa_.load_redform;
% pprior = options_gsa_.pprior;
% ilog = options_gsa_.logtrans_redform;
% threshold = options_gsa_.threshold_redform;
% ksstat = options_gsa_.ksstat_redform;
% alpha2 = options_gsa_.alpha2_redform;
function redform_map(dirname,options_gsa_,M_,estim_params_,options_,bayestopt_,oo_)
% redform_map(dirname,options_gsa_,M_,estim_params_,options_,bayestopt_,oo_)
% Inputs:
% - dirname [string] name of the output directory
% - options_gsa_ [structure] GSA options_
% - M_ [structure] describing the model
% - estim_params_ [structure] characterizing parameters to be estimated
% - options_ [structure] describing the options
% - bayestopt_ [structure] describing the priors
% - oo_ [structure] storing the results
%
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
@ -33,23 +31,16 @@ function redform_map(dirname,options_gsa_)
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
global M_ oo_ estim_params_ options_ bayestopt_
% options_gsa_ = options_.opt_gsa;
anamendo = options_gsa_.namendo;
anamlagendo = options_gsa_.namlagendo;
anamexo = options_gsa_.namexo;
anamendo_tex = options_gsa_.namendo_tex;
anamlagendo_tex = options_gsa_.namlagendo_tex;
anamexo_tex = options_gsa_.namexo_tex;
iload = options_gsa_.load_redform;
pprior = options_gsa_.pprior;
ilog = options_gsa_.logtrans_redform;
threshold = options_gsa_.threshold_redform;
% ksstat = options_gsa_.ksstat_redform;
alpha2 = options_gsa_.alpha2_redform;
alpha2=0;
pvalue_ks = options_gsa_.ksstat_redform;
pvalue_corr = options_gsa_.alpha2_redform;
np = estim_params_.np;
nshock = estim_params_.nvx + estim_params_.nvn + estim_params_.ncx + estim_params_.ncn;
@ -57,11 +48,11 @@ pnames=cell(np,1);
pnames_tex=cell(np,1);
for jj=1:np
if options_.TeX
[param_name_temp, param_name_tex_temp]= get_the_name(nshock+jj,options_.TeX,M_,estim_params_,options_);
pnames_tex{jj,1} = strrep(param_name_tex_temp,'$','');
[param_name_temp, param_name_tex_temp]= get_the_name(nshock+jj,options_.TeX,M_,estim_params_,options_.varobs);
pnames_tex{jj,1} = param_name_tex_temp;
pnames{jj,1} = param_name_temp;
else
param_name_temp = get_the_name(nshock+jj,options_.TeX,M_,estim_params_,options_);
param_name_temp = get_the_name(nshock+jj,options_.TeX,M_,estim_params_,options_.varobs);
pnames{jj,1} = param_name_temp;
end
end
@ -93,14 +84,14 @@ end
options_mcf.fname_ = M_.fname;
options_mcf.OutputDirectoryName = adir;
if ~exist('T')
stab_map_(dirname,options_gsa_);
if ~exist('T','var')
stab_map_(dirname,options_gsa_,M_,oo_,options_,bayestopt_,estim_params_);
if pprior
load([dirname,filesep,M_.fname,'_prior'],'T');
else
load([dirname,filesep,M_.fname,'_mc'],'T');
end
if ~exist('T')
if ~exist('T','var')
disp('The model is too large!')
disp('Reduced form mapping stopped!')
return
@ -109,8 +100,6 @@ end
if isempty(dir(adir))
mkdir(adir)
end
adir0=pwd;
%cd(adir)
nspred=size(T,2)-M_.exo_nbr;
x0=lpmat(istable,:);
@ -121,7 +110,7 @@ else
xx0=lpmat0(istable,:);
nshocks=size(xx0,2);
end
[kn, np]=size(x0);
[~, np]=size(x0);
offset = length(bayestopt_.pshape)-np;
if options_gsa_.prior_range
pshape=5*(ones(np,1));
@ -144,23 +133,22 @@ options_map.pshape = pshape;
options_map.pd = pd;
nsok = length(find(M_.lead_lag_incidence(M_.maximum_lag,:)));
lpmat=[];
lpmat0=[];
js=0;
for j = 1:length(anamendo)
namendo = anamendo{j};
namendo_tex = anamendo_tex{j};
iendo = strmatch(namendo, M_.endo_names(oo_.dr.order_var), 'exact');
ifig = 0;
iplo = 0;
for jx = 1:length(anamexo)
namexo = anamexo{jx};
namexo_tex = anamexo_tex{jx};
iexo=strmatch(namexo, M_.exo_names, 'exact');
skipline()
disp(['[', namendo,' vs ',namexo,']'])
if ~isempty(iexo)
%y0=squeeze(T(iendo,iexo+nspred,istable));
y0=squeeze(T(iendo,iexo+nspred,:));
if (max(y0)-min(y0))>1.e-10
if mod(iplo,9)==0 && isempty(threshold) && ~options_.nograph
@ -196,7 +184,6 @@ for j = 1:length(anamendo)
hf=dyn_figure(options_.nodisplay,'name',['Reduced Form Mapping (Monte Carlo Filtering): ',namendo,' vs ', namexo]);
hc = cumplot(y0);
a=axis; delete(hc);
% hist(mat_moment{ij}),
x1val=max(threshold(1),a(1));
x2val=min(threshold(2),a(2));
hp = patch([x1val x2val x2val x1val],a([3 3 4 4]),'b');
@ -205,9 +192,13 @@ for j = 1:length(anamendo)
hc = cumplot(y0);
set(hc,'color','k','linewidth',2)
hold off,
title([namendo,' vs ', namexo ' - threshold [' num2str(threshold(1)) ' ' num2str(threshold(2)) ']'],'interpreter','none')
if options_.TeX
title([namendo_tex,' vs ', namexo_tex ' - threshold [' num2str(threshold(1)) ' ' num2str(threshold(2)) ']'],'interpreter','latex')
else
title([namendo,' vs ', namexo ' - threshold [' num2str(threshold(1)) ' ' num2str(threshold(2)) ']'],'interpreter','none')
end
dyn_saveas(hf,[xdir,filesep, fname_ '_' type '_' namendo,'_vs_', namexo],options_.nodisplay,options_.graph_format);
create_TeX_loader(options_,[xdir,filesep, fname_ '_' type '_' namendo,'_vs_', namexo],['Reduced Form Mapping (Monte Carlo Filtering): ',strrep(namendo,'_','\_'),' vs ', strrep(namexo,'_','\_')],[type '_' namendo,'_vs_', namexo])
create_TeX_loader(options_,[xdir,filesep, fname_ '_' type '_' namendo,'_vs_', namexo],['Reduced Form Mapping (Monte Carlo Filtering): ',namendo_tex,' vs ', namexo_tex],[type '_' namendo,'_vs_', namexo])
end
si(:,js) = NaN(np,1);
delete([xdir, '/*threshold*.*'])
@ -219,19 +210,23 @@ for j = 1:length(anamendo)
options_mcf.amcf_title = atitle;
options_mcf.beha_title = 'inside threshold';
options_mcf.nobeha_title = 'outside threshold';
if options_.TeX
options_mcf.beha_title_latex = 'inside threshold';
options_mcf.nobeha_title_latex = 'outside threshold';
end
options_mcf.title = atitle0;
options_mcf.OutputDirectoryName = xdir;
if ~isempty(iy) && ~isempty(iyc)
fprintf(['%4.1f%% of the ',type,' support matches ',atitle0,'\n'],length(iy)/length(y0)*100)
icheck = mcf_analysis(x0, iy, iyc, options_mcf, options_);
icheck = mcf_analysis(x0, iy, iyc, options_mcf, M_, options_, bayestopt_, estim_params_);
lpmat=x0(iy,:);
if nshocks
lpmat0=xx0(iy,:);
end
istable=[1:length(iy)];
istable=1:length(iy);
save([xdir,filesep, fname_ '_' type '_' namendo,'_vs_', namexo '_threshold' ],'lpmat','lpmat0','istable','y0','x0','xx0','iy','iyc')
lpmat=[]; lpmat0=[]; istable=[];
lpmat0=[];
if length(iy)<=10 || length(iyc)<=10
icheck = []; % do the generic plot in any case
end
@ -255,12 +250,10 @@ for j = 1:length(anamendo)
end
atitle0=['Monte Carlo Filtering for ',namendo,' vs ', namexo];
options_mcf.title = atitle0;
indmcf = redform_mcf(y0, x0, options_mcf, options_);
redform_mcf(y0, x0, options_mcf, options_, M_.fname, bayestopt_.name, estim_params_);
end
end
else
[yy, xdir] = log_trans_(y0,xdir0);
atitle0=['Reduced Form Mapping (ANOVA) for log-transformed ',namendo,' vs ', namexo];
aname=[type '_' namendo '_vs_' namexo];
atitle=[type ' Reduced Form Mapping (ANOVA): Parameter(s) driving ',namendo,' vs ',namexo];
@ -276,24 +269,31 @@ for j = 1:length(anamendo)
figure(hh_fig)
subplot(3,3,iplo),
if ilog
[saso, iso] = sort(-silog(:,js));
[~, iso] = sort(-silog(:,js));
bar([silog(iso(1:min(np,10)),js)])
logflag='log';
else
[saso, iso] = sort(-si(:,js));
[~, iso] = sort(-si(:,js));
bar(si(iso(1:min(np,10)),js))
logflag='';
end
%set(gca,'xticklabel',pnames(iso(1:min(np,10)),:),'fontsize',8)
set(gca,'xticklabel',' ','fontsize',10)
set(gca,'xlim',[0.5 10.5])
for ip=1:min(np,10)
text(ip,-0.02,deblank(pnames(iso(ip),:)),'rotation',90,'HorizontalAlignment','right','interpreter','none')
if options_.TeX
text(ip,-0.02,deblank(pnames_tex(iso(ip),:)),'rotation',90,'HorizontalAlignment','right','interpreter','latex')
else
text(ip,-0.02,deblank(pnames(iso(ip),:)),'rotation',90,'HorizontalAlignment','right','interpreter','none')
end
end
if options_.TeX
title([logflag,' ',namendo_tex,' vs ',namexo_tex],'interpreter','none')
else
title([logflag,' ',namendo,' vs ',namexo],'interpreter','none')
end
title([logflag,' ',namendo,' vs ',namexo],'interpreter','none')
if iplo==9
dyn_saveas(hh_fig,[dirname,filesep,M_.fname,'_redform_', namendo,'_vs_shocks_',logflag,num2str(ifig)],options_.nodisplay,options_.graph_format);
create_TeX_loader(options_,[dirname,filesep,M_.fname,'_redform_', namendo,'_vs_shocks_',logflag,num2str(ifig)],[logflag,' ',strrep(namendo,'_','\_'),' vs ',strrep(namexo,'_','\_')],['redform_', namendo,'_vs_shocks_',logflag,num2str(ifig)],1)
create_TeX_loader(options_,[dirname,filesep,M_.fname,'_redform_', namendo,'_vs_shocks_',logflag,num2str(ifig)],[logflag,' ',namendo_tex,' vs ',namexo_tex],['redform_', namendo,'_vs_shocks_',logflag,num2str(ifig)],1)
end
end
@ -310,12 +310,12 @@ for j = 1:length(anamendo)
iplo=0;
for je=1:length(anamlagendo)
namlagendo = anamlagendo{je};
namlagendo_tex = anamlagendo_tex{je};
ilagendo=strmatch(namlagendo, M_.endo_names(oo_.dr.order_var(M_.nstatic+1:M_.nstatic+nsok)), 'exact');
skipline()
disp(['[', namendo,' vs lagged ',namlagendo,']'])
if ~isempty(ilagendo)
%y0=squeeze(T(iendo,ilagendo,istable));
y0=squeeze(T(iendo,ilagendo,:));
if (max(y0)-min(y0))>1.e-10
if mod(iplo,9)==0 && isempty(threshold) && ~options_.nograph
@ -331,9 +331,9 @@ for j = 1:length(anamendo)
if isempty(dir(xdir0))
mkdir(xdir0)
end
atitle0=['Reduced Form Mapping (ANOVA) for ',namendo,' vs ', namlagendo];
aname=[type '_' namendo '_vs_' namlagendo];
atitle=[type ' Reduced Form Mapping (ANOVA): Parameter(s) driving ',namendo,' vs ',namlagendo];
atitle0=['Reduced Form Mapping (ANOVA) for ',namendo,' vs lagged', namlagendo];
aname=[type '_' namendo '_vs_lag_' namlagendo];
atitle=[type ' Reduced Form Mapping (ANOVA): Parameter(s) driving ',namendo,' vs lagged',namlagendo];
options_map.amap_name = aname;
options_map.amap_title = atitle;
options_map.figtitle = atitle0;
@ -351,7 +351,6 @@ for j = 1:length(anamendo)
hf=dyn_figure(options_.nodisplay,'name',['Reduced Form Mapping (Monte Carlo Filtering): ',namendo,' vs lagged ', namlagendo]);
hc = cumplot(y0);
a=axis; delete(hc);
% hist(mat_moment{ij}),
x1val=max(threshold(1),a(1));
x2val=min(threshold(2),a(2));
hp = patch([x1val x2val x2val x1val],a([3 3 4 4]),'b');
@ -359,10 +358,14 @@ for j = 1:length(anamendo)
hold all,
hc = cumplot(y0);
set(hc,'color','k','linewidth',2)
hold off,
title([namendo,' vs lagged ', namlagendo ' - threshold [' num2str(threshold(1)) ' ' num2str(threshold(2)) ']'],'interpreter','none')
hold off
if options_.TeX
title([namendo_tex,' vs lagged ', namlagendo_tex ' - threshold [' num2str(threshold(1)) ' ' num2str(threshold(2)) ']'],'interpreter','latex')
else
title([namendo,' vs lagged ', namlagendo ' - threshold [' num2str(threshold(1)) ' ' num2str(threshold(2)) ']'],'interpreter','none')
end
dyn_saveas(hf,[xdir,filesep, fname_ '_' type '_' namendo,'_vs_', namlagendo],options_.nodisplay,options_.graph_format);
create_TeX_loader(options_,[xdir,filesep, fname_ '_' type '_' namendo,'_vs_', namlagendo],['Reduced Form Mapping (Monte Carlo Filtering): ',strrep(namendo,'_','\_'),' vs lagged ', strrep(namlagendo,'_','\_')],[type '_' namendo,'_vs_', namlagendo],1)
create_TeX_loader(options_,[xdir,filesep, fname_ '_' type '_' namendo,'_vs_', namlagendo],['Reduced Form Mapping (Monte Carlo Filtering): ',namendo_tex,' vs lagged ', namlagendo_tex],[type '_' namendo,'_vs_', namlagendo],1)
end
delete([xdir, '/*threshold*.*'])
@ -374,24 +377,28 @@ for j = 1:length(anamendo)
options_mcf.amcf_title = atitle;
options_mcf.beha_title = 'inside threshold';
options_mcf.nobeha_title = 'outside threshold';
if options_.TeX
options_mcf.beha_title_latex = 'inside threshold';
options_mcf.nobeha_title_latex = 'outside threshold';
end
options_mcf.title = atitle0;
options_mcf.OutputDirectoryName = xdir;
if ~isempty(iy) && ~isempty(iyc)
fprintf(['%4.1f%% of the ',type,' support matches ',atitle0,'\n'],length(iy)/length(y0)*100)
icheck = mcf_analysis(x0, iy, iyc, options_mcf, options_);
icheck = mcf_analysis(x0, iy, iyc, options_mcf, M_, options_, bayestopt_, estim_params_);
lpmat=x0(iy,:);
if nshocks
lpmat0=xx0(iy,:);
end
istable=[1:length(iy)];
istable=1:length(iy);
save([xdir,filesep, fname_ '_' type '_' namendo,'_vs_', namlagendo '_threshold' ],'lpmat','lpmat0','istable','y0','x0','xx0','iy','iyc')
lpmat=[]; lpmat0=[]; istable=[];
if length(iy)<=10 || length(iyc)<=10,
lpmat0=[];
if length(iy)<=10 || length(iyc)<=10
icheck = []; % do the generic plot in any case
end
else
icheck = [];
end
@ -412,11 +419,10 @@ for j = 1:length(anamendo)
end
atitle0=['Monte Carlo Filtering for ',namendo,' vs ', namlagendo];
options_mcf.title = atitle0;
indmcf = redform_mcf(y0, x0, options_mcf, options_);
redform_mcf(y0, x0, options_mcf, options_, M_.fname, bayestopt_.name, estim_params_);
end
end
else
[yy, xdir] = log_trans_(y0,xdir0);
atitle0=['Reduced Form Mapping (ANOVA) for log-transformed ',namendo,' vs ', namlagendo];
aname=[type '_' namendo '_vs_' namlagendo];
atitle=[type ' Reduced Form Mapping (ANOVA): Parameter(s) driving ',namendo,' vs ',namlagendo];
@ -432,24 +438,27 @@ for j = 1:length(anamendo)
figure(hh_fig),
subplot(3,3,iplo),
if ilog
[saso, iso] = sort(-silog(:,js));
[~, iso] = sort(-silog(:,js));
bar([silog(iso(1:min(np,10)),js)])
logflag='log';
else
[saso, iso] = sort(-si(:,js));
[~, iso] = sort(-si(:,js));
bar(si(iso(1:min(np,10)),js))
logflag='';
end
%set(gca,'xticklabel',pnames(iso(1:min(np,10)),:),'fontsize',8)
set(gca,'xticklabel',' ','fontsize',10)
set(gca,'xlim',[0.5 10.5])
for ip=1:min(np,10)
text(ip,-0.02,deblank(pnames(iso(ip),:)),'rotation',90,'HorizontalAlignment','right','interpreter','none')
if options_.TeX
text(ip,-0.02,deblank(pnames_tex(iso(ip),:)),'rotation',90,'HorizontalAlignment','right','interpreter','latex')
else
text(ip,-0.02,deblank(pnames(iso(ip),:)),'rotation',90,'HorizontalAlignment','right','interpreter','none')
end
end
title([logflag,' ',namendo,' vs ',namlagendo,'(-1)'],'interpreter','none')
if iplo==9
dyn_saveas(hh_fig,[dirname,filesep,M_.fname,'_redform_', namendo,'_vs_lags_',logflag,num2str(ifig)],options_.nodisplay,options_.graph_format);
create_TeX_loader(options_,[dirname,filesep,M_.fname,'_redform_', namendo,'_vs_lags_',logflag,num2str(ifig)],[logflag,' ',strrep(namendo,'_','\_'),' vs ',strrep(namlagendo,'_','\_'),'(-1)'],['redform_', namendo,'_vs_lags_',logflag,':',num2str(ifig)],1)
create_TeX_loader(options_,[dirname,filesep,M_.fname,'_redform_', namendo,'_vs_lags_',logflag,num2str(ifig)],[logflag,' ',namendo_tex,' vs ',namlagendo_tex,'(-1)'],['redform_', namendo,'_vs_lags_',logflag,':',num2str(ifig)],1)
end
end
@ -460,48 +469,37 @@ for j = 1:length(anamendo)
end
if iplo<9 && iplo>0 && ifig && ~options_.nograph
dyn_saveas(hh_fig,[dirname,filesep,M_.fname,'_redform_', namendo,'_vs_lags_',logflag,num2str(ifig)],options_.nodisplay,options_.graph_format);
create_TeX_loader(options_,[dirname,filesep,M_.fname,'_redform_', namendo,'_vs_lags_',logflag,num2str(ifig)],[logflag,' ',strrep(namendo,'_','\_'),' vs ',strrep(namlagendo,'_','\_'),'(-1)'],['redform_', namendo,'_vs_lags_',logflag,':',num2str(ifig)],options_.figures.textwidth*min(iplo/3,1));
create_TeX_loader(options_,[dirname,filesep,M_.fname,'_redform_', namendo,'_vs_lags_',logflag,num2str(ifig)],[logflag,' ',namendo_tex,' vs ',namlagendo_tex,'(-1)'],['redform_', namendo,'_vs_lags_',logflag,':',num2str(ifig)],options_.figures.textwidth*min(iplo/3,1));
end
end
if isempty(threshold) && ~options_.nograph
hh_fig=dyn_figure(options_.nodisplay,'name','Reduced Form GSA');
if ilog==0
hh_fig=dyn_figure(options_.nodisplay,'name','Reduced Form GSA'); %bar(si)
% boxplot(si','whis',10,'symbol','r.')
myboxplot(si',[],'.',[],10)
xlabel(' ')
set(gca,'xticklabel',' ','fontsize',10,'xtick',[1:np])
set(gca,'xlim',[0.5 np+0.5])
set(gca,'ylim',[0 1])
set(gca,'position',[0.13 0.2 0.775 0.7])
for ip=1:np
text(ip,-0.02,deblank(pnames(ip,:)),'rotation',90,'HorizontalAlignment','right','interpreter','none')
end
else
myboxplot(silog',[],'.',[],10)
end
xlabel(' ')
set(gca,'xticklabel',' ','fontsize',10,'xtick',1:np)
set(gca,'xlim',[0.5 np+0.5])
set(gca,'ylim',[0 1])
set(gca,'position',[0.13 0.2 0.775 0.7])
for ip=1:np
text(ip,-0.02,deblank(pnames(ip,:)),'rotation',90,'HorizontalAlignment','right','interpreter','none')
end
if ilog==0
title('Reduced form GSA')
dyn_saveas(hh_fig,[dirname,filesep,M_.fname,'_redform_gsa'],options_.nodisplay,options_.graph_format);
create_TeX_loader(options_,[dirname,filesep,M_.fname,'_redform_gsa'],'Reduced Form GSA','redform_gsa')
else
hh_fig=dyn_figure(options_.nodisplay,'name','Reduced Form GSA'); %bar(silog)
% boxplot(silog','whis',10,'symbol','r.')
myboxplot(silog',[],'.',[],10)
set(gca,'xticklabel',' ','fontsize',10,'xtick',[1:np])
xlabel(' ')
set(gca,'xlim',[0.5 np+0.5])
set(gca,'ylim',[0 1])
set(gca,'position',[0.13 0.2 0.775 0.7])
for ip=1:np
text(ip,-0.02,deblank(pnames(ip,:)),'rotation',90,'HorizontalAlignment','right','interpreter','none')
end
title('Reduced form GSA - Log-transformed elements')
dyn_saveas(hh_fig,[dirname,filesep,M_.fname,'_redform_gsa_log'],options_.nodisplay,options_.graph_format);
create_TeX_loader(options_,[dirname,filesep,M_.fname,'_redform_gsa_log'],'Reduced form GSA - Log-transformed elements','redform_gsa_log')
end
end
function si = redform_private(x0, y0, options_map, options_)
np=size(x0,2);
x00=x0;
ilog = options_map.log_trans;
@ -531,14 +529,12 @@ if iload==0
nest=max(50,nrun/2);
nest=min(250,nest);
nfit=min(1000,nrun);
% dotheplots = (nfit<=nest);
% gsa_ = gsa_sdp(y0(1:nest), x0(1:nest,:), 2, [],[-1 -1 -1 -1 -1 0],[],0,[fname,'_est'], pnames);
[ys,is] = sort(y0);
[~,is] = sort(y0);
istep = ceil(nrun/nest);
if istep>1
iest = is(floor(istep/2):istep:end);
nest = length(iest);
irest = is(setdiff([1:nrun],[floor(istep/2):istep:nrun]));
irest = is(setdiff(1:nrun,floor(istep/2):istep:nrun));
istep = ceil(length(irest)/(nfit-nest));
ifit = union(iest, irest(1:istep:end));
else
@ -550,22 +546,27 @@ if iload==0
ifit = union(ifit, irest(end));
end
nfit=length(ifit);
% ifit = union(iest, irest(randperm(nrun-nest,nfit-nest)));
% ifit = iest;
% nfit=nest;
ipred = setdiff([1:nrun],ifit);
ipred = setdiff(1:nrun,ifit);
if ilog
[y1, tmp, isig, lam] = log_trans_(y0(iest));
[~, ~, isig, lam] = log_trans_(y0(iest));
y1 = log(y0*isig+lam);
end
if ~options_.nograph
hh_fig=dyn_figure(options_.nodisplay,'name',options_map.figtitle);
subplot(221)
if ilog
hist(y1,30)
if isoctave
hist(y1,30)
else
histogram(y1,30)
end
else
hist(y0,30)
if isoctave
hist(y0,30)
else
histogram(y0,30)
end
end
title(options_map.title,'interpreter','none')
subplot(222)
@ -582,15 +583,7 @@ if iload==0
if ilog
[gsa22, gsa1, gsax] = ss_anova_log(y1(iest), x0(iest,:), isig, lam, gsa0);
end
% if (gsa1.out.bic-gsa0.out.bic) < 10,
% y00=y0;
% gsa00=gsa0;
% gsa0=gsa1;
% y0=y1;
% ilog=1;
% end
if nfit>nest
% gsa_ = gsa_sdp(y0(1:nfit), x0(1:nfit,:), -2, gsa_.nvr*nest^3/nfit^3,[-1 -1 -1 -1 -1 0],[],0,fname, pnames);
nvr = gsa0.nvr*nest^3/nfit^3;
nvr(gsa0.stat<2) = gsa0.nvr(gsa0.stat<2)*nest^5/nfit^5;
gsa_ = ss_anova(y0(ifit), x0(ifit,:), 1, 0, 2, nvr);
@ -601,33 +594,6 @@ if iload==0
nvrx = gsax.nvr*nest^3/nfit^3;
nvrx(gsax.stat<2) = gsax.nvr(gsax.stat<2)*nest^5/nfit^5;
[gsa22, gsa1, gsax] = ss_anova_log(y1(ifit), x0(ifit,:), isig, lam, gsa0, [nvr1' nvrx']);
% gsa1 = ss_anova(y1(ifit), x0(ifit,:), 1, 0, 2, nvr);
% gsa2=gsa1;
% gsa2.y = gsa0.y;
% gsa2.fit = (exp(gsa1.fit)-lam)*isig;
% gsa2.f0 = mean(gsa2.fit);
% gsa2.out.SSE = sum((gsa2.fit-gsa2.y).^2);
% gsa2.out.bic = gsa2.out.bic-nest*log(gsa1.out.SSE)+nest*log(gsa2.out.SSE);
% gsa2.r2 = 1-cov(gsa2.fit-gsa2.y)/cov(gsa2.y);
% for j=1:np,
% gsa2.fs(:,j) = exp(gsa1.fs(:,j)).*mean(exp(gsa1.fit-gsa1.f(:,j)))*isig-lam*isig-gsa2.f0;
% gsa2.f(:,j) = exp(gsa1.f(:,j)).*mean(exp(gsa1.fit-gsa1.f(:,j)))*isig-lam*isig-gsa2.f0;
% gsa2.si(j) = var(gsa2.f(:,j))/var(gsa2.y);
% end
% nvr = gsax.nvr*nest^3/nfit^3;
% nvr(gsax.stat<2) = gsax.nvr(gsax.stat<2)*nest^5/nfit^5;
% gsax = ss_anova([gsa2.y-gsa2.fit], x0(ifit,:), 1, 0, 2, nvr);
% gsa22=gsa2;
% gsa22.fit = gsa2.fit+gsax.fit;
% gsa22.f0 = mean(gsa22.fit);
% gsa22.out.SSE = sum((gsa22.fit-gsa22.y).^2);
% gsa22.out.bic = nest*log(gsa22.out.SSE/nest) + (gsax.out.df+gsa2.out.df-1)*log(nest);
% gsa22.r2 = 1-sum((gsa22.fit-gsa22.y).^2)/sum((gsa22.y-mean(gsa22.y)).^2);
% for j=1:np,
% gsa22.fs(:,j) = gsa2.fs(:,j)+gsax.fs(:,j);
% gsa22.f(:,j) = gsa2.f(:,j)+gsax.f(:,j);
% gsa22.si(j) = var(gsa22.f(:,j))/var(gsa22.y);
% end
gsa_ = gsa22;
end
else
@ -638,26 +604,18 @@ if iload==0
end
end
save([fname,'_map.mat'],'gsa_')
[sidum, iii]=sort(-gsa_.si);
[~, iii]=sort(-gsa_.si);
gsa_.x0=x00(ifit,:);
if ~options_.nograph
hmap=gsa_sdp_plot(gsa_,[fname '_map'],pnames,iii(1:min(12,np)));
set(hmap,'name',options_map.amap_title);
end
gsa_.x0=x0(ifit,:);
% copyfile([fname,'_est.mat'],[fname,'.mat'])
if ~options_.nograph
figure(hh_fig);
subplot(223),
plot(y0(ifit),[gsa_.fit y0(ifit)],'.'),
r2 = gsa_.r2;
% if ilog,
% plot(y00(ifit),[log_trans_(gsa_.fit,'',isig,lam) y00(ifit)],'.'),
% r2 = 1 - cov(log_trans_(gsa_.fit,'',isig,lam)-y00(ifit))/cov(y00(ifit));
% else
% plot(y0(ifit),[gsa_.fit y0(ifit)],'.'),
% r2 = gsa_.r2;
% end
title(['Learning sample fit - R2=' num2str(r2,2)],'interpreter','none')
if nfit<nrun
if ilog
@ -680,8 +638,6 @@ if iload==0
end
end
else
% gsa_ = gsa_sdp_dyn(y0, x0, 0, [],[],[],0,fname, pnames);
% gsa_ = gsa_sdp(y0, x0, 0, [],[],[],0,fname, pnames);
load([fname,'_map.mat'],'gsa_')
if ~options_.nograph
yf = ss_anova_fcast(x0, gsa_);
@ -690,10 +646,8 @@ else
title([namy,' vs ', namx,' pred'],'interpreter','none')
dyn_saveas(hh_fig,[fname '_pred'],options_.nodisplay,options_.graph_format);
create_TeX_loader(options_,[fname '_pred'],options_map.title,[namy,' vs ', namx,' pred'])
end
end
% si = gsa_.multivariate.si;
si = gsa_.si;
return
@ -730,6 +684,8 @@ else
end
gsa2 = log2level_map(gsa1, isig, lam);
if nargin >=5 && ~isempty(gsa0)
nvr2=NaN(np,1);
nvr0=NaN(np,1);
for j=1:np
nvr2(j) = var(diff(gsa2.fs(:,j),2));
nvr0(j) = var(diff(gsa0.fs(:,j),2));
@ -760,26 +716,24 @@ end
return
function indmcf = redform_mcf(y0, x0, options_mcf, options_)
function indmcf = redform_mcf(y0, x0, options_mcf, options_, fname, parnames, estim_params_)
hh_fig=dyn_figure(options_.nodisplay,'name',options_mcf.amcf_title);
[post_mean, post_median, post_var, hpd_interval, post_deciles, ...
density] = posterior_moments(y0,1,0.9);
[~, ~, ~, ~, post_deciles] = posterior_moments(y0,1,0.9);
post_deciles = [-inf; post_deciles; inf];
for jt=1:10
indy{jt}=find( (y0>post_deciles(jt)) & (y0<=post_deciles(jt+1)));
leg{jt}=[int2str(jt) '-dec'];
end
[proba, dproba] = stab_map_1(x0, indy{1}, indy{end}, [],0);
[proba] = stab_map_1(x0, indy{1}, indy{end}, [], fname, options_, parnames, estim_params_,0);
indmcf=find(proba<options_mcf.pvalue_ks);
if isempty(indmcf)
[tmp,jtmp] = sort(proba,2,'ascend');
[~,jtmp] = sort(proba,1,'ascend');
indmcf = jtmp(1);
% indmcf = jtmp(1:min(2,length(proba)));
end
[tmp,jtmp] = sort(proba(indmcf),2,'ascend');
[~,jtmp] = sort(proba(indmcf),1,'ascend');
indmcf = indmcf(jtmp);
nbr_par = length(indmcf);
nrow=ceil(sqrt(nbr_par+1));
@ -797,7 +751,11 @@ for jx=1:nbr_par
set(h,'color', cmap(jt,:), 'linewidth', 2)
hold all
end
title(options_mcf.param_names(indmcf(jx),:),'interpreter','none')
if options_.TeX
title(options_mcf.param_names_tex(indmcf(jx),:),'interpreter','latex')
else
title(options_mcf.param_names(indmcf(jx),:),'interpreter','none')
end
end
hleg = legend(leg);
aa=get(hleg,'Position');

90
matlab/gsa/redform_screen.m
View File

@ -1,10 +1,14 @@
function redform_screen(dirname, options_gsa_)
%function redform_map(dirname, options_gsa_)
% inputs (from opt_gsa structure
% anamendo = options_gsa_.namendo;
% anamlagendo = options_gsa_.namlagendo;
% anamexo = options_gsa_.namexo;
% iload = options_gsa_.load_redform;
function redform_screen(dirname, options_gsa_, estim_params_, M_, dr, options_, bayestopt_)
% redform_screen(dirname, options_gsa_, estim_params_, M_, dr, options_, bayestopt_)
% Conduct reduced form screening
% Inputs:
% - dirname [string] name of the output directory
% - options_gsa_ [structure] GSA options_
% - estim_params [structure] describing the estimated parameters
% - M_ [structure] describing the model
% - dr [structure] decision rules
% - options_ [structure] describing the options
% - bayestopt_ [structure] describing the priors
%
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
@ -28,17 +32,21 @@ function redform_screen(dirname, options_gsa_)
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
global M_ oo_ estim_params_ options_ bayestopt_
% options_gsa_ = options_.opt_gsa;
anamendo = options_gsa_.namendo;
anamlagendo = options_gsa_.namlagendo;
anamexo = options_gsa_.namexo;
iload = options_gsa_.load_redform;
anamendo_tex = options_gsa_.namendo_tex;
anamlagendo_tex = options_gsa_.namlagendo_tex;
anamexo_tex = options_gsa_.namexo_tex;
nliv = options_gsa_.morris_nliv;
pnames = M_.param_names(estim_params_.param_vals(:,1));
if options_.TeX
for par_iter=1:size(estim_params_.param_vals(:,1),1)
[~,tex_names{par_iter,1}]=get_the_name(estim_params_.param_vals(par_iter,1),options_.TeX, M_, estim_params_, options_.varobs);
end
end
if nargin==0
dirname='';
end
@ -54,35 +62,45 @@ nsok = length(find(M_.lead_lag_incidence(M_.maximum_lag,:)));
js=0;
for j=1:size(anamendo,1)
namendo = deblank(anamendo(j,:));
iendo = strmatch(namendo, M_.endo_names(oo_.dr.order_var), 'exact');
namendo = anamendo{j,:};
namendo_tex = anamendo_tex{j,:};
iendo = strmatch(namendo, M_.endo_names(dr.order_var), 'exact');
iplo=0;
ifig=0;
for jx=1:size(anamexo,1)
namexo = deblank(anamexo(jx,:));
namexo = anamexo{jx};
namexo_tex = anamexo_tex{jx};
iexo = strmatch(namexo, M_.exo_names, 'exact');
if ~isempty(iexo)
y0=teff(T(iendo,iexo+nspred,:), kn, istable);
if ~isempty(y0)
if mod(iplo,9)==0
ifig = ifig+1;
hh_fig = dyn_figure(options_.nodisplay, 'name', [namendo,[' vs. shocks '], int2str(ifig)]);
hh_fig = dyn_figure(options_.nodisplay, 'name', [namendo,' vs. shocks ', int2str(ifig)]);
iplo = 0;
end
iplo = iplo+1;
js = js+1;
subplot(3, 3, iplo)
[SAmeas, SAMorris] = Morris_Measure_Groups(np+nshock, [lpmat0 lpmat], y0, nliv);
[~, SAMorris] = Morris_Measure_Groups(np+nshock, [lpmat0 lpmat], y0, nliv);
SAM = squeeze(SAMorris(nshock+1:end,1));
SA(:,js) = SAM./(max(SAM)+eps);
[saso, iso] = sort(-SA(:,js));
[~, iso] = sort(-SA(:,js));
bar(SA(iso(1:min(np,10)),js))
set(gca,'xticklabel',' ','fontsize',10)
set(gca,'xlim',[0.5 10.5])
for ip=1:min(np,10)
text(ip,-0.02,pnames(iso(ip)),'rotation',90,'HorizontalAlignment','right','interpreter','none')
if options_.TeX
text(ip,-0.02,tex_names(iso(ip)),'rotation',90,'HorizontalAlignment','right','interpreter','latex')
else
text(ip,-0.02,pnames(iso(ip)),'rotation',90,'HorizontalAlignment','right','interpreter','none')
end
end
if options_.TeX
title([namendo_tex,' vs. ',namexo_tex],'interpreter','latex')
else
title([namendo,' vs. ',namexo],'interpreter','none')
end
title([namendo,' vs. ',namexo],'interpreter','none')
if iplo==9
dyn_saveas(hh_fig,[dirname,'/',M_.fname,'_', namendo,'_vs_shock_',num2str(ifig)],options_.nodisplay,options_.graph_format);
create_TeX_loader(options_,[dirname,'/',M_.fname,'_', namendo,'_vs_shock_',num2str(ifig)],ifig,[namendo,' vs. shocks ',int2str(ifig)],[namendo,'_vs_shock'],1)
@ -99,8 +117,9 @@ for j=1:size(anamendo,1)
iplo=0;
ifig=0;
for je=1:size(anamlagendo,1)
namlagendo=deblank(anamlagendo(je,:));
ilagendo=strmatch(namlagendo, M_.endo_names(oo_.dr.order_var(M_.nstatic+1:M_.nstatic+nsok)), 'exact');
namlagendo=anamlagendo{je};
namlagendo_tex=anamlagendo_tex{je};
ilagendo=strmatch(namlagendo, M_.endo_names(dr.order_var(M_.nstatic+1:M_.nstatic+nsok)), 'exact');
if ~isempty(ilagendo)
y0=teff(T(iendo,ilagendo,:),kn,istable);
@ -113,19 +132,26 @@ for j=1:size(anamendo,1)
iplo=iplo+1;
js=js+1;
subplot(3,3,iplo),
[SAmeas, SAMorris] = Morris_Measure_Groups(np+nshock, [lpmat0 lpmat], y0,nliv);
[~, SAMorris] = Morris_Measure_Groups(np+nshock, [lpmat0 lpmat], y0,nliv);
SAM = squeeze(SAMorris(nshock+1:end,1));
SA(:,js)=SAM./(max(SAM)+eps);
[saso, iso] = sort(-SA(:,js));
[~, iso] = sort(-SA(:,js));
bar(SA(iso(1:min(np,10)),js))
%set(gca,'xticklabel',pnames(iso(1:min(np,10)),:),'fontsize',8)
set(gca,'xticklabel',' ','fontsize',10)
set(gca,'xlim',[0.5 10.5])
for ip=1:min(np,10)
text(ip,-0.02,deblank(pnames(iso(ip),:)),'rotation',90,'HorizontalAlignment','right','interpreter','none')
if options_.TeX
text(ip,-0.02,tex_names(iso(ip)),'rotation',90,'HorizontalAlignment','right','interpreter','latex')
else
text(ip,-0.02,pnames{iso(ip)},'rotation',90,'HorizontalAlignment','right','interpreter','none')
end
end
title([namendo,' vs. ',namlagendo,'(-1)'],'interpreter','none')
if options_.TeX
title([namendo_tex,' vs. ',namlagendo_tex,'(-1)'],'interpreter','latex')
else
title([namendo,' vs. ',namlagendo,'(-1)'],'interpreter','none')
end
if iplo==9
dyn_saveas(hh_fig,[dirname,'/',M_.fname,'_', namendo,'_vs_lags_',num2str(ifig)],options_.nodisplay,options_.graph_format);
create_TeX_loader(options_,[dirname,'/',M_.fname,'_', namendo,'_vs_lags_',num2str(ifig)],ifig,[namendo,' vs. lagged endogenous ',int2str(ifig)],[namendo,'_vs_lags'],1)
@ -140,15 +166,17 @@ for j=1:size(anamendo,1)
end
hh_fig=dyn_figure(options_.nodisplay,'Name','Reduced form screening');
%bar(SA)
% boxplot(SA','whis',10,'symbol','r.')
myboxplot(SA',[],'.',[],10)
set(gca,'xticklabel',' ','fontsize',10,'xtick',[1:np])
set(gca,'xticklabel',' ','fontsize',10,'xtick',1:np)
set(gca,'xlim',[0.5 np+0.5])
set(gca,'ylim',[0 1])
set(gca,'position',[0.13 0.2 0.775 0.7])
for ip=1:np
text(ip,-0.02,deblank(pnames(ip,:)),'rotation',90,'HorizontalAlignment','right','interpreter','none')
if options_.TeX
text(ip,-0.02,tex_names(ip),'rotation',90,'HorizontalAlignment','right','interpreter','latex')
else
text(ip,-0.02,pnames{ip},'rotation',90,'HorizontalAlignment','right','interpreter','none')
end
end
xlabel(' ')
ylabel('Elementary Effects')

13
matlab/gsa/scatter_analysis.m
View File

@ -1,4 +1,6 @@
function indmcf = scatter_analysis(lpmat, xdata, options_scatter, options_)
function scatter_analysis(lpmat, xdata, options_scatter, options_)
% scatter_analysis(lpmat, xdata, options_scatter, options_)
% Plot scatter plot analysis
%
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
@ -6,7 +8,7 @@ function indmcf = scatter_analysis(lpmat, xdata, options_scatter, options_)
%
% Copyright © 2017 European Commission
% Copyright © 2017 Dynare Team
% Copyright © 2017-2023 Dynare Team
%
% This file is part of Dynare.
%
@ -34,7 +36,6 @@ if options_.TeX
end
amcf_name = options_scatter.amcf_name;
amcf_title = options_scatter.amcf_title;
title = options_scatter.title;
fname_ = options_scatter.fname_;
xparam1=[];
if isfield(options_scatter,'xparam1')
@ -48,5 +49,9 @@ if ~options_.nograph
if ~isempty(xparam1)
xx=xparam1;
end
scatter_plots(lpmat, xdata, param_names, '.', [fname_, '_', amcf_name], OutputDirectoryName, amcf_title, xx, options_)
if options_.TeX
scatter_plots(lpmat, xdata, param_names_tex, '.', [fname_, '_', amcf_name], OutputDirectoryName, amcf_title, xx, options_)
else
scatter_plots(lpmat, xdata, param_names, '.', [fname_, '_', amcf_name], OutputDirectoryName, amcf_title, xx, options_)
end
end

40
matlab/gsa/scatter_callback.m
View File

@ -1,40 +0,0 @@
function scatter_callback(K, type)
%
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
% marco.ratto@ec.europa.eu
%
% Copyright © 2017 European Commission
% Copyright © 2018 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 <https://www.gnu.org/licenses/>.
global oo_ M_ options_ bayestopt_ estim_params_
x=get(gcf,'userdata');
r2=x{1};
x=x{2};
xparam1=x(K,:)';
switch type
case 'save'
save(['my_params_' int2str(K)],'xparam1')
case 'eval'
disp('Evaluating smoother ...')
[oo_, M_]=evaluate_smoother(xparam1, M_.endo_names, M_, oo_, options_, bayestopt_, estim_params_);
end

50
matlab/gsa/scatter_mcf.m
View File

@ -1,5 +1,5 @@
function scatter_mcf(X,Y,vnames,plotsymbol, fnam, dirname, figtitle, xparam1, options_, beha_name, non_beha_name)
% scatter_mcf(X,Y,vnames,plotsymbol, fnam, dirname, figtitle, xparam1, options_, beha_name, non_beha_name)
function scatter_mcf(X,Y,vnames,plotsymbol, fnam, dirname, figtitle, xparam1, options_, beha_name, non_beha_name, beha_name_latex, non_beha_name_latex)
% scatter_mcf(X,Y,vnames,plotsymbol, fnam, dirname, figtitle, xparam1, options_, beha_name, non_beha_name, beha_name_latex, non_beha_name_latex)
%
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
@ -38,12 +38,6 @@ function scatter_mcf(X,Y,vnames,plotsymbol, fnam, dirname, figtitle, xparam1, o
Z=[X;Y];
[n,p] = size(X);
% X = X - ones(n,1)*min(Z);
% X = X ./ (ones(n,1)*max(Z));
[n,p] = size(Y);
% Y = Y - ones(n,1)*min(Z);
% Y = Y ./ (ones(n,1)*max(Z));
[n,p] = size(Z);
clear Z;
@ -53,8 +47,10 @@ if nargin >=3
end
if nargin<4 || isempty(plotsymbol)
if n*p<100, plotsymbol = 'o';
else plotsymbol = '.';
if n*p<100
plotsymbol = 'o';
else
plotsymbol = '.';
end
end
@ -83,7 +79,7 @@ end
figtitle_tex=strrep(figtitle,'_','\_');
fig_nam_=[fnam];
fig_nam_=fnam;
if ~nograph
hh_fig=dyn_figure(options_.nodisplay,'name',figtitle);
end
@ -101,7 +97,6 @@ for i = 1:p
h = axes('position',[fL(i),fL(p+1-j),ffl,ffl]);
if i==j
h1=cumplot(X(:,j));
% set(h1,'color',[0 0 1], 'linestyle','--','LineWidth',1.5)
set(h1,'color',[0 0 1],'LineWidth',1.5)
hold on,
h2=cumplot(Y(:,j));
@ -126,10 +121,10 @@ for i = 1:p
plot(X(:,i),X(:,j),[plotsymbol,'b'])
end
if ~isempty(xparam1)
hold on, plot(xparam1(i),xparam1(j),'s','MarkerFaceColor',[0 0.75 0],'MarkerEdgeColor',[0 0.75 0])
hold on
plot(xparam1(i),xparam1(j),'s','MarkerFaceColor',[0 0.75 0],'MarkerEdgeColor',[0 0.75 0])
end
hold off;
% axis([-0.1 1.1 -0.1 1.1])
if i<p
set(gca,'YTickLabel',[],'YTick',[]);
else
@ -144,16 +139,26 @@ for i = 1:p
end
if i==1
if nflag == 1
ylabel(vnames(j,:),'Rotation',45, ...
'HorizontalAlignment','right','VerticalAlignment','middle');
if options_.TeX
ylabel(vnames(j,:),'Rotation',45, ...
'HorizontalAlignment','right','VerticalAlignment','middle','Interpreter','latex');
else
ylabel(vnames(j,:),'Rotation',45, ...
'HorizontalAlignment','right','VerticalAlignment','middle','Interpreter','none');
end
else
ylabel([num2str(j),' '],'Rotation',90)
end
end
if j==1
if nflag == 1
title(vnames(i,:),'Rotation',45, ...
'HorizontalAlignment','left','VerticalAlignment','bottom')
if options_.TeX
title(vnames(i,:),'Rotation',45, ...
'HorizontalAlignment','left','VerticalAlignment','bottom','Interpreter','latex')
else
title(vnames(i,:),'Rotation',45, ...
'HorizontalAlignment','left','VerticalAlignment','bottom','Interpreter','none')
end
else
title(num2str(i))
end
@ -162,8 +167,13 @@ for i = 1:p
end
end
if ~isoctave
annotation('textbox', [0.1,0,0.35,0.05],'String', beha_name,'Color','Blue','horizontalalignment','center','interpreter','none');
annotation('textbox', [0.55,0,0.35,0.05],'String', non_beha_name,'Color','Red','horizontalalignment','center','interpreter','none');
if options_.TeX
annotation('textbox', [0.1,0,0.35,0.05],'String', beha_name_latex,'Color','Blue','horizontalalignment','center','interpreter','latex');
annotation('textbox', [0.55,0,0.35,0.05],'String', non_beha_name_latex,'Color','Red','horizontalalignment','center','interpreter','latex');
else
annotation('textbox', [0.1,0,0.35,0.05],'String', beha_name,'Color','Blue','horizontalalignment','center','interpreter','none');
annotation('textbox', [0.55,0,0.35,0.05],'String', non_beha_name,'Color','Red','horizontalalignment','center','interpreter','none');
end
end
if ~nograph

89
matlab/gsa/scatter_plots.m
View File

@ -1,9 +1,21 @@
function scatter_plots(X,xp,vnames,plotsymbol, fnam, dirname, figtitle, xparam1, options_)
% scatter_plots(X,xp,vnames,plotsymbol, fnam, dirname, figtitle, xparam1, options_)
% Pairwise scatter plots of the columns of x and y after Monte Carlo filtering
% Inputs:
% - X [double] nxk matrix with columns containing behavioural sample
% - xp [double] mxk matrix with columns containing non-behavioural sample
% - vnames [char] vector of variable names (default = numeric labels 1,2,3 etc.)
% - plotsymbol [char] plt symbol (default = '.' for npts > 100, 'o' for npts < 100
% - fnam [char] figure name
% - dirname [char] directory name
% - figtitle [char] figure title
% - xparam1 [double] parameter vector
% - options_ [struct] option structure
%
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
% marco.ratto@ec.europa.eu
%
% Copyright © 2017 European Commission
% Copyright © 2017-2023 Dynare Team
@ -23,23 +35,7 @@ function scatter_plots(X,xp,vnames,plotsymbol, fnam, dirname, figtitle, xparam1,
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
% PURPOSE: Pairwise scatter plots of the columns of x and y after
% Monte Carlo filtering
%---------------------------------------------------
% USAGE: scatter_mcf(x,y,vnames,pltsym,diagon)
% or scatter_mcf(x,y) which relies on defaults
% where:
% x = an nxk matrix with columns containing behavioural sample
% y = an mxk matrix with columns containing non-behavioural sample
% vnames = a vector of variable names
% (default = numeric labels 1,2,3 etc.)
% pltsym = a plt symbol
% (default = '.' for npts > 100, 'o' for npts < 100
[n,p] = size(X);
% X = X - ones(n,1)*min(Z);
% X = X ./ (ones(n,1)*max(Z));
nflag = 0;
if nargin >=3
@ -47,8 +43,10 @@ if nargin >=3
end
if nargin<4 || isempty(plotsymbol)
if n*p<100, plotsymbol = 'o';
else plotsymbol = '.';
if n*p<100
plotsymbol = 'o';
else
plotsymbol = '.';
end
end
@ -71,7 +69,7 @@ end
figtitle_tex=strrep(figtitle,'_','\_');
fig_nam_=[fnam];
fig_nam_=fnam;
hh_fig=dyn_figure(options_.nodisplay,'name',figtitle);
set(hh_fig,'userdata',{X,xp})
@ -90,7 +88,6 @@ for i = 1:p
if i==j
h1=cumplot(X(:,j));
set(h,'Tag','cumplot')
% set(h1,'color',[0 0 1], 'linestyle','--','LineWidth',1.5)
set(h1,'color',[0 0 1],'LineWidth',1.5)
if ~isempty(xparam1)
hold on, plot(xparam1([j j]),[0 1],'k--')
@ -102,39 +99,13 @@ for i = 1:p
end
set(gca,'YTickLabel',[],'YTick',[]);
else
if j>i
plot(X(:,i),X(:,j),[plotsymbol,'b'])
else
plot(X(:,i),X(:,j),[plotsymbol,'b'])
end
plot(X(:,i),X(:,j),[plotsymbol,'b'])
set(h,'Tag','scatter')
%%
if ~isoctave
% Define a context menu; it is not attached to anything
hcmenu = uicontextmenu('Callback','pick','Tag','Run viewer');
% Define callbacks for context menu
% items that change linestyle
hcb1 = 'scatter_callback';
% hcb2 = ['set(gco,''LineStyle'','':'')'];
% hcb3 = ['set(gco,''LineStyle'',''-'')'];
% % Define the context menu items and install their callbacks
item1 = uimenu(hcmenu,'Label','save','Callback',hcb1,'Tag','save params');
item2 = uimenu(hcmenu,'Label','eval','Callback',hcb1,'Tag','eval params');
% item3 = uimenu(hcmenu,'Label','solid','Callback',hcb3);
% Locate line objects
hlines = findall(h,'Type','line');
% Attach the context menu to each line
for line = 1:length(hlines)
set(hlines(line),'uicontextmenu',hcmenu)
end
end
%%
if ~isempty(xparam1)
hold on, plot(xparam1(i),xparam1(j),'s','MarkerFaceColor',[0 0.75 0],'MarkerEdgeColor',[0 0.75 0])
end
hold off;
% axis([-0.1 1.1 -0.1 1.1])
if i<p
set(gca,'YTickLabel',[],'YTick',[]);
else
@ -149,16 +120,26 @@ for i = 1:p
end
if i==1
if nflag == 1
ylabel(vnames(j,:),'Rotation',45,'interpreter','none', ...
'HorizontalAlignment','right','VerticalAlignment','middle');
if options_.TeX
ylabel(vnames(j,:),'Rotation',45,'interpreter','latex', ...
'HorizontalAlignment','right','VerticalAlignment','middle');
else
ylabel(vnames(j,:),'Rotation',45,'interpreter','none', ...
'HorizontalAlignment','right','VerticalAlignment','middle');
end
else
ylabel([num2str(j),' '],'Rotation',90)
end
end
if j==1
if nflag == 1
title(vnames(i,:),'interpreter','none','Rotation',45, ...
'HorizontalAlignment','left','VerticalAlignment','bottom')
if options_.TeX
title(vnames(i,:),'interpreter','latex','Rotation',45, ...
'HorizontalAlignment','left','VerticalAlignment','bottom')
else
title(vnames(i,:),'interpreter','none','Rotation',45, ...
'HorizontalAlignment','left','VerticalAlignment','bottom')
end
else
title(num2str(i))
end
@ -166,10 +147,6 @@ for i = 1:p
drawnow
end
end
% if ~isoctave
% annotation('textbox', [0.1,0,0.35,0.05],'String', beha_name,'Color','Blue','horizontalalignment','center','interpreter','none');
% annotation('textbox', [0.55,0,0.35,0.05],'String', non_beha_name,'Color','Red','horizontalalignment','center','interpreter','none');
% end
if ~nograph
dyn_saveas(hh_fig,[dirname,filesep,fig_nam_],options_.nodisplay,options_.graph_format);

16
matlab/gsa/set_shocks_param.m
View File

@ -1,8 +1,16 @@
function set_shocks_param(xparam1)
% function set_shocks_param(xparam1)
function M_=set_shocks_param(M_,estim_params_,xparam1)
% function M_=set_shocks_param(M_,estim_params_,xparam1)
% Set the structural and measurement error variances and covariances
% Inputs
% - M_ [structure] Matlab's structure describing the model
% - estim_params_ [structure] characterizing parameters to be estimated
% - xparam1 [double] parameter vector
% Outputs:
% - M_ [structure] Matlab's structure describing the model
%
% Notes: closely follows set_all_parameters.m
% Copyright © 2012-2017 Dynare Team
% Copyright © 2012-2023 Dynare Team
%
% This file is part of Dynare.
%
@ -19,8 +27,6 @@ function set_shocks_param(xparam1)
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
global estim_params_ M_
nvx = estim_params_.nvx;
ncx = estim_params_.ncx;
nvn = estim_params_.nvn;

189
matlab/gsa/stab_map_.m
View File

@ -1,9 +1,22 @@
function x0 = stab_map_(OutputDirectoryName,opt_gsa)
% x0 = stab_map_(OutputDirectoryName,opt_gsa)
function x0 = stab_map_(OutputDirectoryName,opt_gsa,M_,oo_,options_,bayestopt_,estim_params_)
% x0 = stab_map_(OutputDirectoryName,opt_gsa,M_,oo_,options_,bayestopt_,estim_params_)
% Mapping of stability regions in the prior ranges applying
% Monte Carlo filtering techniques.
%
% INPUTS (from opt_gsa structure)
% Inputs
% - OutputDirectoryName [string] name of the output directory
% - opt_gsa [structure] GSA options structure
% - M_ [structure] Matlab's structure describing the model
% - oo_ [structure] Matlab's structure describing the results
% - options_ [structure] Matlab's structure describing the current options
% - bayestopt_ [structure] describing the priors
% - estim_params_ [structure] characterizing parameters to be estimated
%
% Outputs:
% - x0 one parameter vector for which the model is stable.
%
%
% Inputs from opt_gsa structure
% Nsam = MC sample size
% fload = 0 to run new MC; 1 to load prevoiusly generated analysis
% alpha2 = significance level for bivariate sensitivity analysis
@ -14,8 +27,6 @@ function x0 = stab_map_(OutputDirectoryName,opt_gsa)
% _prior.mat file
% = 0: sample from posterior ranges: sample saved in
% _mc.mat file
% OUTPUT:
% x0: one parameter vector for which the model is stable.
%
% GRAPHS
% 1) Pdf's of marginal distributions under the stability (dotted
@ -50,11 +61,6 @@ function x0 = stab_map_(OutputDirectoryName,opt_gsa)
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
%global bayestopt_ estim_params_ dr_ options_ ys_ fname_
global bayestopt_ estim_params_ options_ oo_ M_
% opt_gsa=options_.opt_gsa;
Nsam = opt_gsa.Nsam;
fload = opt_gsa.load_stab;
alpha2 = opt_gsa.alpha2_stab;
@ -82,6 +88,7 @@ nshock = nshock + estim_params_.ncn;
lpmat0=zeros(Nsam,0);
xparam1=[];
%% prepare prior bounds
[~,~,~,lb,ub,~] = set_prior(estim_params_,M_,options_); %Prepare bounds
if ~isempty(bayestopt_) && any(bayestopt_.pshape > 0)
% Set prior bounds
@ -107,15 +114,16 @@ options_mcf.pvalue_ks = pvalue_ks;
options_mcf.pvalue_corr = pvalue_corr;
options_mcf.alpha2 = alpha2;
%% get LaTeX names
name=cell(np,1);
name_tex=cell(np,1);
for jj=1:np
if options_.TeX
[param_name_temp, param_name_tex_temp]= get_the_name(nshock+jj,options_.TeX,M_,estim_params_,options_);
name_tex{jj,1} = strrep(param_name_tex_temp,'$','');
[param_name_temp, param_name_tex_temp]= get_the_name(nshock+jj,options_.TeX,M_,estim_params_,options_.varobs);
name_tex{jj,1} = param_name_tex_temp;
name{jj,1} = param_name_temp;
else
param_name_temp = get_the_name(nshock+jj,options_.TeX,M_,estim_params_,options_);
param_name_temp = get_the_name(nshock+jj,options_.TeX,M_,estim_params_,options_.varobs);
name{jj,1} = param_name_temp;
end
end
@ -128,19 +136,18 @@ options_mcf.fname_ = fname_;
options_mcf.OutputDirectoryName = OutputDirectoryName;
options_mcf.xparam1 = [];
opt=options_;
options_.periods=0;
options_.nomoments=1;
options_.irf=0;
options_.noprint=1;
if fload==0
if fload==0 %run new MC
if isfield(dr_,'ghx')
egg=zeros(length(dr_.eigval),Nsam);
end
yys=zeros(length(dr_.ys),Nsam);
if opt_gsa.morris == 1
[lpmat, OutFact] = Sampling_Function_2(nliv, np+nshock, ntra, ones(np+nshock, 1), zeros(np+nshock,1), []);
[lpmat] = Sampling_Function_2(nliv, np+nshock, ntra, ones(np+nshock, 1), zeros(np+nshock,1), []);
lpmat = lpmat.*(nliv-1)/nliv+1/nliv/2;
Nsam=size(lpmat,1);
lpmat0 = lpmat(:,1:nshock);
@ -158,10 +165,9 @@ if fload==0
for j=1:np
lpmat(:,j) = randperm(Nsam)'./(Nsam+1); %latin hypercube
end
end
end
dummy=prior_draw_gsa(1);
prior_draw_gsa(M_,bayestopt_,options_,estim_params_,1); %initialize
if pprior
for j=1:nshock
if opt_gsa.morris~=1
@ -178,16 +184,16 @@ if fload==0
lpmat(:,j)=lpmat(:,j).*(upper_bound-lower_bound)+lower_bound;
end
else
xx=prior_draw_gsa(0,[lpmat0 lpmat]);
xx=prior_draw_gsa(M_,bayestopt_,options_,estim_params_,0,[lpmat0 lpmat]);
lpmat0=xx(:,1:nshock);
lpmat=xx(:,nshock+1:end);
clear xx;
end
else
else %posterior analysis
if neighborhood_width>0 && isempty(options_.mode_file)
xparam1 = get_all_parameters(estim_params_,M_);
else
eval(['load ' options_.mode_file '.mat;']);
load([options_.mode_file '.mat'],'hh','xparam1');
end
if neighborhood_width>0
for j=1:nshock
@ -216,8 +222,8 @@ if fload==0
for j=1:Nsam*2
lnprior(j) = any(lp(j,:)'<=bounds.lb | lp(j,:)'>=bounds.ub);
end
ireal=[1:2*Nsam];
ireal=ireal(find(lnprior==0));
ireal=1:2*Nsam;
ireal=ireal(lnprior==0);
lp=lp(ireal,:);
Nsam=min(Nsam, length(ireal));
lpmat0=lp(1:Nsam,1:nshock);
@ -227,9 +233,9 @@ if fload==0
end
%
h = dyn_waitbar(0,'Please wait...');
istable=[1:Nsam];
istable=1:Nsam;
jstab=0;
iunstable=[1:Nsam];
iunstable=1:Nsam;
iindeterm=zeros(1,Nsam);
iwrong=zeros(1,Nsam);
inorestriction=zeros(1,Nsam);
@ -237,12 +243,11 @@ if fload==0
infox=zeros(Nsam,1);
for j=1:Nsam
M_ = set_all_parameters([lpmat0(j,:) lpmat(j,:)]',estim_params_,M_);
%try stoch_simul([]);
try
if ~ isempty(options_.endogenous_prior_restrictions.moment)
[Tt,Rr,SteadyState,info,oo_.dr,M_.params] = dynare_resolve(M_,options_,oo_.dr,oo_.steady_state,oo_.exo_steady_state,oo_.exo_det_steady_state);
if ~isempty(options_.endogenous_prior_restrictions.moment)
[Tt,Rr,~,info,oo_.dr,M_.params] = dynare_resolve(M_,options_,oo_.dr,oo_.steady_state,oo_.exo_steady_state,oo_.exo_det_steady_state);
else
[Tt,Rr,SteadyState,info,oo_.dr,M_.params] = dynare_resolve(M_,options_,oo_.dr,oo_.steady_state,oo_.exo_steady_state,oo_.exo_det_steady_state,'restrict');
[Tt,Rr,~,info,oo_.dr,M_.params] = dynare_resolve(M_,options_,oo_.dr,oo_.steady_state,oo_.exo_steady_state,oo_.exo_det_steady_state,'restrict');
end
infox(j,1)=info(1);
if infox(j,1)==0 && ~exist('T','var')
@ -250,8 +255,7 @@ if fload==0
if prepSA
try
T=zeros(size(dr_.ghx,1),size(dr_.ghx,2)+size(dr_.ghu,2),Nsam);
catch
ME = lasterror();
catch ME
if strcmp('MATLAB:nomem',ME.identifier)
prepSA=0;
disp('The model is too large for storing state space matrices ...')
@ -324,8 +328,9 @@ if fload==0
end
ys_=real(dr_.ys);
yys(:,j) = ys_;
ys_=yys(:,1);
dyn_waitbar(j/Nsam,h,['MC iteration ',int2str(j),'/',int2str(Nsam)])
if mod(j,3)
dyn_waitbar(j/Nsam,h,['MC iteration ',int2str(j),'/',int2str(Nsam)])
end
end
dyn_waitbar_close(h);
if prepSA && jstab
@ -333,13 +338,13 @@ if fload==0
else
T=[];
end
istable=istable(find(istable)); % stable params ignoring restrictions
irestriction=irestriction(find(irestriction)); % stable params & restrictions OK
inorestriction=inorestriction(find(inorestriction)); % stable params violating restrictions
iunstable=iunstable(find(iunstable)); % violation of BK & restrictions & solution could not be found (whatever goes wrong)
iindeterm=iindeterm(find(iindeterm)); % indeterminacy
iwrong=iwrong(find(iwrong)); % dynare could not find solution
ixun=iunstable(find(~ismember(iunstable,[iindeterm,iwrong,inorestriction]))); % explosive roots
istable=istable(istable~=0); % stable params ignoring restrictions
irestriction=irestriction(irestriction~=0); % stable params & restrictions OK
inorestriction=inorestriction(inorestriction~=0); % stable params violating restrictions
iunstable=iunstable(iunstable~=0); % violation of BK & restrictions & solution could not be found (whatever goes wrong)
iindeterm=iindeterm(iindeterm~=0); % indeterminacy
iwrong=iwrong(iwrong~=0); % dynare could not find solution
ixun=iunstable(~ismember(iunstable,[iindeterm,iwrong,inorestriction])); % explosive roots
bkpprior.pshape=bayestopt_.pshape;
bkpprior.p1=bayestopt_.p1;
@ -356,8 +361,7 @@ if fload==0
'bkpprior','lpmat','lpmat0','irestriction','iunstable','istable','iindeterm','iwrong','ixun', ...
'egg','yys','T','nspred','nboth','nfwrd','infox')
end
else
else %~pprior
if ~prepSA
save([OutputDirectoryName filesep fname_ '_mc.mat'], ...
'lpmat','lpmat0','irestriction','iunstable','istable','iindeterm','iwrong','ixun', ...
@ -368,19 +372,18 @@ if fload==0
'egg','yys','T','nspred','nboth','nfwrd','infox')
end
end
else
else %load old run
if pprior
filetoload=[OutputDirectoryName filesep fname_ '_prior.mat'];
else
filetoload=[OutputDirectoryName filesep fname_ '_mc.mat'];
end
load(filetoload,'lpmat','lpmat0','irestriction','iunstable','istable','iindeterm','iwrong','ixun','egg','yys','nspred','nboth','nfwrd','infox')
load(filetoload,'lpmat','lpmat0','irestriction','iunstable','istable','iindeterm','iwrong','ixun','infox')
Nsam = size(lpmat,1);
if pprior==0 && ~isempty(options_.mode_file)
eval(['load ' options_.mode_file '.mat;']);
load([options_.mode_file '.mat'],'xparam1');
end
if prepSA && isempty(strmatch('T',who('-file', filetoload),'exact'))
h = dyn_waitbar(0,'Please wait...');
options_.periods=0;
@ -392,30 +395,24 @@ else
yys=NaN(length(ys_),ntrans);
for j=1:ntrans
M_.params(estim_params_.param_vals(:,1)) = lpmat(istable(j),:)';
%stoch_simul([]);
[Tt,Rr,SteadyState,info,oo_.dr,M_.params] = dynare_resolve(M_,options_,oo_.dr,oo_.steady_state,oo_.exo_steady_state,oo_.exo_det_steady_state,'restrict');
[~,~,~,~,oo_.dr,M_.params] = dynare_resolve(M_,options_,oo_.dr,oo_.steady_state,oo_.exo_steady_state,oo_.exo_det_steady_state,'restrict');
if ~exist('T','var')
T=zeros(size(dr_.ghx,1),size(dr_.ghx,2)+size(dr_.ghu,2),ntrans);
end
dr_ = oo_.dr;
T(:,:,j) = [dr_.ghx dr_.ghu];
if ~exist('nspred','var')
nspred = dr_.nspred; %size(dr_.ghx,2);
nboth = dr_.nboth;
nfwrd = dr_.nfwrd;
end
ys_=real(dr_.ys);
yys(:,j) = ys_;
ys_=yys(:,1);
dyn_waitbar(j/ntrans,h,['MC iteration ',int2str(j),'/',int2str(ntrans)])
if mod(j,3)
dyn_waitbar(j/ntrans,h,['MC iteration ',int2str(j),'/',int2str(ntrans)])
end
end
dyn_waitbar_close(h);
save(filetoload,'T','-append')
elseif prepSA
load(filetoload,'T')
end
end
%% display and save output
if pprior
aunstname='prior_unstable'; aunsttitle='Prior StabMap: explosiveness of solution';
aindname='prior_indeterm'; aindtitle='Prior StabMap: Indeterminacy';
@ -435,17 +432,18 @@ delete([OutputDirectoryName,filesep,fname_,'_',aindname,'.*']);
delete([OutputDirectoryName,filesep,fname_,'_',aunstname,'.*']);
delete([OutputDirectoryName,filesep,fname_,'_',awrongname,'.*']);
if length(iunstable)>0 || length(iwrong)>0
fprintf(['%4.1f%% of the prior support gives unique saddle-path solution.\n'],length(istable)/Nsam*100)
fprintf(['%4.1f%% of the prior support gives explosive dynamics.\n'],(length(ixun) )/Nsam*100)
fprintf('\nSensitivity Analysis: Stability mapping:\n')
if ~isempty(iunstable) || ~isempty(iwrong)
fprintf('%4.1f%% of the prior support gives unique saddle-path solution.\n',length(istable)/Nsam*100)
fprintf('%4.1f%% of the prior support gives explosive dynamics.\n',(length(ixun) )/Nsam*100)
if ~isempty(iindeterm)
fprintf(['%4.1f%% of the prior support gives indeterminacy.\n'],length(iindeterm)/Nsam*100)
fprintf('%4.1f%% of the prior support gives indeterminacy.\n',length(iindeterm)/Nsam*100)
end
inorestriction = istable(find(~ismember(istable,irestriction))); % violation of prior restrictions
inorestriction = istable(~ismember(istable,irestriction)); % violation of prior restrictions
if ~isempty(iwrong) || ~isempty(inorestriction)
skipline()
if any(infox==49)
fprintf(['%4.1f%% of the prior support violates prior restrictions.\n'],(length(inorestriction) )/Nsam*100)
fprintf('%4.1f%% of the prior support violates prior restrictions.\n',(length(inorestriction) )/Nsam*100)
end
if ~isempty(iwrong)
skipline()
@ -486,50 +484,66 @@ if length(iunstable)>0 || length(iwrong)>0
end
skipline()
if length(iunstable)<Nsam || length(istable)>1
itot = [1:Nsam];
isolve = itot(find(~ismember(itot,iwrong))); % dynare could find a solution
itot = 1:Nsam;
isolve = itot(~ismember(itot,iwrong)); % dynare could find a solution
% Blanchard Kahn
if neighborhood_width
options_mcf.xparam1 = xparam1(nshock+1:end);
end
itmp = itot(find(~ismember(itot,istable)));
itmp = itot(~ismember(itot,istable));
options_mcf.amcf_name = asname;
options_mcf.amcf_title = atitle;
options_mcf.beha_title = 'unique Stable Saddle-Path';
options_mcf.nobeha_title = 'NO unique Stable Saddle-Path';
if options_.TeX
options_mcf.beha_title_latex = 'unique Stable Saddle-Path';
options_mcf.nobeha_title_latex = 'NO unique Stable Saddle-Path';
end
options_mcf.title = 'unique solution';
mcf_analysis(lpmat, istable, itmp, options_mcf, options_);
mcf_analysis(lpmat, istable, itmp, options_mcf, M_, options_, bayestopt_, estim_params_);
if ~isempty(iindeterm)
itmp = isolve(find(~ismember(isolve,iindeterm)));
itmp = isolve(~ismember(isolve,iindeterm));
options_mcf.amcf_name = aindname;
options_mcf.amcf_title = aindtitle;
options_mcf.beha_title = 'NO indeterminacy';
options_mcf.nobeha_title = 'indeterminacy';
if options_.TeX
options_mcf.beha_title_latex = 'NO indeterminacy';
options_mcf.nobeha_title_latex = 'indeterminacy';
end
options_mcf.title = 'indeterminacy';
mcf_analysis(lpmat, itmp, iindeterm, options_mcf, options_);
mcf_analysis(lpmat, itmp, iindeterm, options_mcf, M_, options_, bayestopt_, estim_params_);
end
if ~isempty(ixun)
itmp = isolve(find(~ismember(isolve,ixun)));
itmp = isolve(~ismember(isolve,ixun));
options_mcf.amcf_name = aunstname;
options_mcf.amcf_title = aunsttitle;
options_mcf.beha_title = 'NO explosive solution';
options_mcf.nobeha_title = 'explosive solution';
if options_.TeX
options_mcf.beha_title_latex = 'NO explosive solution';
options_mcf.nobeha_title_latex = 'explosive solution';
end
options_mcf.title = 'instability';
mcf_analysis(lpmat, itmp, ixun, options_mcf, options_);
mcf_analysis(lpmat, itmp, ixun, options_mcf, M_, options_, bayestopt_, estim_params_);
end
inorestriction = istable(find(~ismember(istable,irestriction))); % violation of prior restrictions
iwrong = iwrong(find(~ismember(iwrong,inorestriction))); % what went wrong beyond prior restrictions
inorestriction = istable(~ismember(istable,irestriction)); % violation of prior restrictions
iwrong = iwrong(~ismember(iwrong,inorestriction)); % what went wrong beyond prior restrictions
if ~isempty(iwrong)
itmp = itot(find(~ismember(itot,iwrong)));
itmp = itot(~ismember(itot,iwrong));
options_mcf.amcf_name = awrongname;
options_mcf.amcf_title = awrongtitle;
options_mcf.beha_title = 'NO inability to find a solution';
options_mcf.nobeha_title = 'inability to find a solution';
if options_.TeX
options_mcf.beha_title_latex = 'NO inability to find a solution';
options_mcf.nobeha_title_latex = 'inability to find a solution';
end
options_mcf.title = 'inability to find a solution';
mcf_analysis(lpmat, itmp, iwrong, options_mcf, options_);
mcf_analysis(lpmat, itmp, iwrong, options_mcf, M_, options_, bayestopt_, estim_params_);
end
if ~isempty(irestriction)
@ -541,11 +555,11 @@ if length(iunstable)>0 || length(iwrong)>0
name_tex=cell(np,1);
for jj=1:np
if options_.TeX
[param_name_temp, param_name_tex_temp]= get_the_name(jj,options_.TeX,M_,estim_params_,options_);
name_tex{jj,1} = strrep(param_name_tex_temp,'$','');
[param_name_temp, param_name_tex_temp]= get_the_name(jj,options_.TeX,M_,estim_params_,options_.varobs);
name_tex{jj,1} = param_name_tex_temp;
name{jj,1} = param_name_temp;
else
param_name_temp = get_the_name(jj,options_.TeX,M_,estim_params_,options_);
param_name_temp = get_the_name(jj,options_.TeX,M_,estim_params_,options_.varobs);
name{jj,1} = param_name_temp;
end
end
@ -557,8 +571,12 @@ if length(iunstable)>0 || length(iwrong)>0
options_mcf.amcf_title = acalibtitle;
options_mcf.beha_title = 'prior IRF/moment calibration';
options_mcf.nobeha_title = 'NO prior IRF/moment calibration';
if options_.TeX
options_mcf.beha_title_latex = 'prior IRF/moment calibration';
options_mcf.nobeha_title_latex = 'NO prior IRF/moment calibration';
end
options_mcf.title = 'prior restrictions';
mcf_analysis([lpmat0 lpmat], irestriction, inorestriction, options_mcf, options_);
mcf_analysis([lpmat0 lpmat], irestriction, inorestriction, options_mcf, M_, options_, bayestopt_, estim_params_);
iok = irestriction(1);
x0 = [lpmat0(iok,:)'; lpmat(iok,:)'];
else
@ -568,7 +586,7 @@ if length(iunstable)>0 || length(iwrong)>0
end
M_ = set_all_parameters(x0,estim_params_,M_);
[oo_.dr,info,M_.params] = resol(0,M_,options_,oo_.dr ,oo_.steady_state, oo_.exo_steady_state, oo_.exo_det_steady_state);
[oo_.dr,~,M_.params] = resol(0,M_,options_,oo_.dr ,oo_.steady_state, oo_.exo_steady_state, oo_.exo_det_steady_state);
else
disp('All parameter values in the specified ranges are not acceptable!')
x0=[];
@ -578,15 +596,8 @@ else
disp('and match prior IRF/moment restriction(s) if any!')
x0=0.5.*(bounds.ub(1:nshock)-bounds.lb(1:nshock))+bounds.lb(1:nshock);
x0 = [x0; lpmat(istable(1),:)'];
end
skipline(1);
xparam1=x0;
save([OutputDirectoryName filesep 'prior_ok.mat'],'xparam1');
options_.periods=opt.periods;
if isfield(opt,'nomoments')
options_.nomoments=opt.nomoments;
end
options_.irf=opt.irf;
options_.noprint=opt.noprint;
save([OutputDirectoryName filesep 'prior_ok.mat'],'xparam1');

53
matlab/gsa/stab_map_1.m
View File

@ -1,16 +1,20 @@
function [proba, dproba] = stab_map_1(lpmat, ibehaviour, inonbehaviour, aname, iplot, ipar, dirname, pcrit, atitle)
%function [proba, dproba] = stab_map_1(lpmat, ibehaviour, inonbehaviour, aname, iplot, ipar, dirname, pcrit)
%
% lpmat = Monte Carlo matrix
% ibehaviour = index of behavioural runs
% inonbehaviour = index of non-behavioural runs
% aname = label of the analysis
% iplot = 1 plot cumulative distributions (default)
% iplot = 0 no plots
% ipar = index array of parameters to plot
% dirname = (OPTIONAL) path of the directory where to save
% (default: current directory)
% pcrit = (OPTIONAL) critical value of the pvalue below which show the plots
function [proba, dproba] = stab_map_1(lpmat, ibehaviour, inonbehaviour, aname, fname_, options_, parnames, estim_params_, iplot, ipar, dirname, pcrit, atitle)
% [proba, dproba] = stab_map_1(lpmat, ibehaviour, inonbehaviour, aname, fname_, options_, parnames, estim_params_, iplot, ipar, dirname, pcrit, atitle)
% Inputs:
% - lpmat [double] Monte Carlo matrix
% - ibehaviour [integer] index of behavioural runs
% - inonbehaviour [integer] index of non-behavioural runs
% - aname [string] label of the analysis
% - fname_ [string] file name
% - options_ [structure] options structure
% - parnames [char] parameter name vector
% - estim_params_ [structure] characterizing parameters to be estimated
% - iplot [boolean] 1 plot cumulative distributions (default)
% 0 no plots
% - ipar [integer] index array of parameters to plot
% - dirname [string] (OPTIONAL) path of the directory where to save
% (default: current directory)
% - pcrit [double] (OPTIONAL) critical value of the pvalue below which show the plots
%
% Plots: dotted lines for BEHAVIOURAL
% solid lines for NON BEHAVIOURAL
@ -38,16 +42,13 @@ function [proba, dproba] = stab_map_1(lpmat, ibehaviour, inonbehaviour, aname, i
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
global estim_params_ bayestopt_ M_ options_
if nargin<5
if nargin<9
iplot=1;
end
fname_ = M_.fname;
if nargin<7
if nargin<11
dirname='';
end
if nargin<9,
if nargin<13
atitle=aname;
end
@ -59,27 +60,28 @@ nshock = nshock + estim_params_.ncn;
npar=size(lpmat,2);
ishock= npar>estim_params_.np;
if nargin<6
if nargin<10
ipar=[];
end
if nargin<8 || isempty(pcrit)
if nargin<12 || isempty(pcrit)
pcrit=1;
end
% Smirnov test for Blanchard;
% Smirnov test for Blanchard
proba=NaN(npar,1);
dproba=NaN(npar,1);
for j=1:npar
[H,P,KSSTAT] = smirnov(lpmat(ibehaviour,j),lpmat(inonbehaviour,j));
[~,P,KSSTAT] = smirnov(lpmat(ibehaviour,j),lpmat(inonbehaviour,j));
proba(j)=P;
dproba(j)=KSSTAT;
end
if isempty(ipar)
% ipar=find(dproba>dcrit);
ipar=find(proba<pcrit);
end
nparplot=length(ipar);
if iplot && ~options_.nograph
lpmat=lpmat(:,ipar);
ftit=bayestopt_.name(ipar+nshock*(1-ishock));
ftit=parnames(ipar+nshock*(1-ishock));
for i=1:ceil(nparplot/12)
hh_fig=dyn_figure(options_.nodisplay,'name',atitle);
@ -94,7 +96,6 @@ if iplot && ~options_.nograph
h=cumplot(lpmat(inonbehaviour,j));
set(h,'color',[0 0 0],'LineWidth',1.5)
end
% title([ftit{j},'. D-stat ', num2str(dproba(ipar(j)),2)],'interpreter','none')
title([ftit{j},'. p-value ', num2str(proba(ipar(j)),2)],'interpreter','none')
end
if nparplot>12

113
matlab/gsa/stab_map_2.m
View File

@ -1,6 +1,20 @@
function indcorr = stab_map_2(x,alpha2, pvalue_crit, fnam, dirname,xparam1,figtitle)
% function stab_map_2(x, alpha2, pvalue, fnam, dirname,xparam1)
function indcorr = stab_map_2(x,alpha2, pvalue_crit, M_,options_,bayestopt_,estim_params_, case_name_plain, case_name_latex, dirname,xparam1,figtitle,fig_caption_latex)
% indcorr = stab_map_2(x,alpha2, pvalue_crit, M_,options_,bayestopt_,estim_params_, fnam, fnam_latex, dirname,xparam1,figtitle,fig_caption_latex)
% Inputs:
% - x
% - alpha2
% - pvalue_crit
% - M_ [structure] Matlab's structure describing the model
% - options_ [structure] Matlab's structure describing the current options
% - bayestopt_ [structure] describing the priors
% - estim_params_ [structure] characterizing parameters to be estimated
% - fnam [string] file name
% - dirnam [string] directory name
% - xparam1 [double] parameter vector
% - figtitle [string] figure title
%
% Output:
% - indcorr
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
% marco.ratto@ec.europa.eu
@ -22,29 +36,30 @@ function indcorr = stab_map_2(x,alpha2, pvalue_crit, fnam, dirname,xparam1,figti
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
%global bayestopt_ estim_params_ dr_ options_ ys_ fname_
global bayestopt_ estim_params_ options_ oo_ M_
npar=size(x,2);
nsam=size(x,1);
ishock= npar>estim_params_.np;
nograph = options_.nograph;
if nargin<4
fnam='';
if nargin<8
case_name_plain='';
end
if nargin<5
if nargin<9
case_name_latex=case_name_plain;
end
if nargin<10
dirname='';
nograph=1;
end
if nargin<6
if nargin<11
xparam1=[];
end
if nargin<7
figtitle=fnam;
if nargin<12
figtitle=case_name_plain;
end
if nargin<13
fig_caption_latex=case_name_latex;
end
ys_ = oo_.dr.ys;
dr_ = oo_.dr;
fname_ = M_.fname;
nshock = estim_params_.nvx;
nshock = nshock + estim_params_.nvn;
@ -53,9 +68,9 @@ nshock = nshock + estim_params_.ncn;
[c0, pvalue] = corrcoef(x);
c00=tril(c0,-1);
fig_nam_=[fname_,'_',fnam,'_corr_'];
fig_nam_tex_table=strrep([fnam,'_corr'],' ','_');
fig_nam_=strrep(fig_nam_,' ','_');
fig_nam_save=[fname_,'_',case_name_plain,'_corr_'];
fig_nam_save=strrep(fig_nam_save,' ','_');
fig_nam_tex_table_save=strrep([case_name_plain,'_corr'],' ','_');
ifig=0;
j2=0;
@ -67,47 +82,43 @@ if ishock==0
else
npar=estim_params_.np+nshock;
end
title_string=['Correlation analysis for ',fnam];
title_string_tex=['Correlation analysis for ',strrep(fnam,'_','\\_')];
title_string=['Correlation analysis for ',case_name_plain];
title_string_tex=['Correlation analysis for ',case_name_latex];
indcorr = [];
entry_iter=1;
for j=1:npar
i2=find(abs(c00(:,j))>alpha2);
if length(i2)>0
if ~isempty(i2)
for jx=1:length(i2)
if pvalue(j,i2(jx))<pvalue_crit
indcorr = [indcorr; [j i2(jx)]];
j2=j2+1;
if ishock
if options_.TeX
[param_name_temp1, param_name_tex_temp1]= get_the_name(j,options_.TeX,M_,estim_params_,options_);
param_name_tex_temp1 = strrep(param_name_tex_temp1,'$','');
[param_name_temp2, param_name_tex_temp2]= get_the_name(i2(jx),options_.TeX,M_,estim_params_,options_);
param_name_tex_temp2 = strrep(param_name_tex_temp2,'$','');
[param_name_temp1, param_name_tex_temp1]= get_the_name(j,options_.TeX,M_,estim_params_,options_.varobs);
[param_name_temp2, param_name_tex_temp2]= get_the_name(i2(jx),options_.TeX,M_,estim_params_,options_.varobs);
tmp_name=(['[',param_name_temp1,',',param_name_temp2,']']);
tmp_name_tex=(['[',param_name_tex_temp1,',',param_name_tex_temp2,']']);
name{entry_iter,1}=tmp_name;
name_tex{entry_iter,1}=tmp_name_tex;
name_tex{entry_iter,1}=strrep(tmp_name_tex,'$',''); %prevent $ inside of expression for table
else
[param_name_temp1]= get_the_name(j,options_.TeX,M_,estim_params_,options_);
[param_name_temp2]= get_the_name(i2(jx),options_.TeX,M_,estim_params_,options_);
[param_name_temp1]= get_the_name(j,options_.TeX,M_,estim_params_,options_.varobs);
[param_name_temp2]= get_the_name(i2(jx),options_.TeX,M_,estim_params_,options_.varobs);
tmp_name=(['[',param_name_temp1,',',param_name_temp2,']']);
name{entry_iter,1}=tmp_name;
end
else
if options_.TeX
[param_name_temp1, param_name_tex_temp1]= get_the_name(j+nshock,options_.TeX,M_,estim_params_,options_);
param_name_tex_temp1 = strrep(param_name_tex_temp1,'$','');
[param_name_temp2, param_name_tex_temp2]= get_the_name(i2(jx)+nshock,options_.TeX,M_,estim_params_,options_);
param_name_tex_temp2 = strrep(param_name_tex_temp2,'$','');
[param_name_temp1, param_name_tex_temp1]= get_the_name(j+nshock,options_.TeX,M_,estim_params_,options_.varobs);
[param_name_temp2, param_name_tex_temp2]= get_the_name(i2(jx)+nshock,options_.TeX,M_,estim_params_,options_.varobs);
tmp_name=(['[',param_name_temp1,',',param_name_temp2,']']);
tmp_name_tex=(['[',param_name_tex_temp1,',',param_name_tex_temp2,']']);
name{entry_iter,1}=tmp_name;
name_tex{entry_iter,1}=tmp_name_tex;
name_tex{entry_iter,1}=strrep(tmp_name_tex,'$',''); %prevent $ inside of expression for table
else
[param_name_temp1]= get_the_name(j+nshock,options_.TeX,M_,estim_params_,options_);
[param_name_temp2]= get_the_name(i2(jx)+nshock,options_.TeX,M_,estim_params_,options_);
[param_name_temp1]= get_the_name(j+nshock,options_.TeX,M_,estim_params_,options_.varobs);
[param_name_temp2]= get_the_name(i2(jx)+nshock,options_.TeX,M_,estim_params_,options_.varobs);
tmp_name=(['[',param_name_temp1,',',param_name_temp2,']']);
name{entry_iter,1}=tmp_name;
end
@ -121,17 +132,10 @@ for j=1:npar
hh_fig=dyn_figure(options_.nodisplay,'name',[figtitle,' sample bivariate projection ', num2str(ifig)]);
end
subplot(3,4,j2-(ifig-1)*12)
% bar(c0(i2,j)),
% set(gca,'xticklabel',bayestopt_.name(i2)),
% set(gca,'xtick',[1:length(i2)])
%plot(stock_par(ixx(nfilt+1:end,i),j),stock_par(ixx(nfilt+1:end,i),i2(jx)),'.k')
%hold on,
plot(x(:,j),x(:,i2(jx)),'.')
if ~isempty(xparam1)
hold on, plot(xparam1(j),xparam1(i2(jx)),'ro')
end
% xlabel(deblank(estim_params_.param_names(j,:)),'interpreter','none'),
% ylabel(deblank(estim_params_.param_names(i2(jx),:)),'interpreter','none'),
if ishock
xlabel(bayestopt_.name{j},'interpreter','none'),
ylabel(bayestopt_.name{i2(jx)},'interpreter','none'),
@ -141,16 +145,16 @@ for j=1:npar
end
title(['cc = ',num2str(c0(i2(jx),j))])
if (mod(j2,12)==0) && j2>0
dyn_saveas(hh_fig,[dirname,filesep,fig_nam_,int2str(ifig)],options_.nodisplay,options_.graph_format);
dyn_saveas(hh_fig,[dirname,filesep,fig_nam_save,int2str(ifig)],options_.nodisplay,options_.graph_format);
if options_.TeX && any(strcmp('eps',cellstr(options_.graph_format)))
fidTeX = fopen([dirname,filesep,fig_nam_,int2str(ifig),'.tex'],'w');
fidTeX = fopen([dirname,filesep,fig_nam_save,int2str(ifig),'.tex'],'w');
fprintf(fidTeX,'%% TeX eps-loader file generated by stab_map_2.m (Dynare).\n');
fprintf(fidTeX,['%% ' datestr(now,0) '\n\n']);
fprintf(fidTeX,'\\begin{figure}[H]\n');
fprintf(fidTeX,'\\centering \n');
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s}\n',strrep([dirname,'/',fig_nam_,int2str(ifig)],'\','/'));
fprintf(fidTeX,'\\caption{%s.}',[figtitle,' sample bivariate projection ', num2str(ifig)]);
fprintf(fidTeX,'\\label{Fig:%s:%u}\n',fig_nam_,ifig);
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s}\n',strrep([dirname,'/',fig_nam_save,int2str(ifig)],'\','/'));
fprintf(fidTeX,'\\caption{%s.}',[fig_caption_latex,' sample bivariate projection ', num2str(ifig)]);
fprintf(fidTeX,'\\label{Fig:%s:%u}\n',fig_nam_save,ifig);
fprintf(fidTeX,'\\end{figure}\n\n');
fprintf(fidTeX,'%% End Of TeX file. \n');
fclose(fidTeX);
@ -162,16 +166,16 @@ for j=1:npar
end
end
if ~nograph && (j==(npar)) && j2>0 && (mod(j2,12)~=0)
dyn_saveas(hh_fig,[dirname,filesep,fig_nam_,int2str(ifig)],options_.nodisplay,options_.graph_format);
dyn_saveas(hh_fig,[dirname,filesep,fig_nam_save,int2str(ifig)],options_.nodisplay,options_.graph_format);
if options_.TeX && any(strcmp('eps',cellstr(options_.graph_format)))
fidTeX = fopen([dirname,filesep,fig_nam_,int2str(ifig),'.tex'],'w');
fidTeX = fopen([dirname,filesep,fig_nam_save,int2str(ifig),'.tex'],'w');
fprintf(fidTeX,'%% TeX eps-loader file generated by stab_map_2.m (Dynare).\n');
fprintf(fidTeX,['%% ' datestr(now,0) '\n\n']);
fprintf(fidTeX,'\\begin{figure}[H]\n');
fprintf(fidTeX,'\\centering \n');
fprintf(fidTeX,'\\includegraphics[width=%2.2f\\textwidth]{%s}\n',options_.figures.textwidth*min((j2-(ifig-1)*12)/3,1),strrep([dirname,'/',fig_nam_,int2str(ifig)],'\','/'));
fprintf(fidTeX,'\\caption{%s.}',[figtitle,' sample bivariate projection ', num2str(ifig)]);
fprintf(fidTeX,'\\label{Fig:%s:%u}\n',fig_nam_,ifig);
fprintf(fidTeX,'\\includegraphics[width=%2.2f\\textwidth]{%s}\n',options_.figures.textwidth*min((j2-(ifig-1)*12)/3,1),strrep([dirname,'/',fig_nam_save,int2str(ifig)],'\','/'));
fprintf(fidTeX,'\\caption{%s.}',[fig_caption_latex,' sample bivariate projection ', num2str(ifig)]);
fprintf(fidTeX,'\\label{Fig:%s:%u}\n',fig_nam_save,ifig);
fprintf(fidTeX,'\\end{figure}\n\n');
fprintf(fidTeX,'%% End Of TeX file. \n');
fclose(fidTeX);
@ -181,7 +185,7 @@ end
if j2==0
skipline();
disp(['No correlation term with pvalue <', num2str(pvalue_crit),' and |corr. coef.| >',num2str(alpha2),' found for ',fnam])
disp(['No correlation term with pvalue <', num2str(pvalue_crit),' and |corr. coef.| >',num2str(alpha2),' found for ',case_name_plain])
else
headers={'Parameters'; 'corrcoef'};
if ~options_.noprint
@ -189,7 +193,6 @@ else
end
dyntable(options_,title_string,headers, name, data_mat, 0, 7, 3);
if options_.TeX
dyn_latex_table(M_, options_, title_string_tex, fig_nam_tex_table, headers, name_tex, data_mat, 0, 7, 3);
dyn_latex_table(M_, options_, title_string_tex, fig_nam_tex_table_save, headers, name_tex, data_mat, 0, 7, 3);
end
end
%close all
end

24
matlab/gsa/stand_.m
View File

@ -1,20 +1,22 @@
function [y, meany, stdy] = stand_(x)
% STAND_ Standardise a matrix by columns
% [y, meany, stdy] = stand_(x)
% Standardise a matrix by columns
%
% [x,my,sy]=stand(y)
%
% y: Time series (column matrix)
% Inputs:
% - x: Time series (column matrix)
%
% x: standardised equivalent of y
% my: Vector of mean values for each column of y
% sy: Vector of standard deviations for each column of y
% - y: standardised equivalent of x
% - meany: Vector of mean values for each column of x
% - stdy: Vector of standard deviations for each column of x
%
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
% marco.ratto@ec.europa.eu
% Copyright © 2012 European Commission
% Copyright © 2012-2017 Dynare Team%
% Copyright © 2012-2023 Dynare Team
% This file is part of Dynare.
%
% Dynare is free software: you can redistribute it and/or modify
@ -34,9 +36,11 @@ if nargin==0
return
end
meany=NaN(size(x,2),1);
stdy=NaN(size(x,2),1);
y=NaN(size(x));
for j=1:size(x,2)
meany(j)=mean(x(find(~isnan(x(:,j))),j));
stdy(j)=std(x(find(~isnan(x(:,j))),j));
meany(j)=mean(x(~isnan(x(:,j)),j));
stdy(j)=std(x(~isnan(x(:,j)),j));
y(:,j)=(x(:,j)-meany(j))./stdy(j);
end
% end of m-file
end

2
matlab/gsa/tcrit.m
View File

@ -150,4 +150,4 @@ if n<=100
t_crit=t_crit(n,ncol);
else
t_crit=t_crit(end,ncol);
end
end

6
matlab/gsa/teff.m
View File

@ -37,15 +37,12 @@ if ndim==3
[ir, ic]=(find( (tmax-tmin)>1.e-8));
j0 = length(ir);
yt=zeros(Nsam, j0);
for j=1:j0
y0=squeeze(T(ir(j),ic(j),:));
%y1=ones(size(lpmat,1),1)*NaN;
y1=ones(Nsam,1)*NaN;
y1(istable,1)=y0;
yt(:,j)=y1;
end
else
tmax=max(T,[],2);
tmin=min(T,[],2);
@ -53,7 +50,4 @@ else
j0 = length(ir);
yt=NaN(Nsam, j0);
yt(istable,:)=T(ir,:)';
end
%clear y0 y1;

42
matlab/gsa/th_moments.m
View File

@ -1,7 +1,22 @@
function [vdec, corr, autocorr, z, zz] = th_moments(dr,var_list)
% [vdec, corr, autocorr, z, zz] = th_moments(dr,var_list)
function [vdec, corr, autocorr, z, zz] = th_moments(dr,options_,M_)
% [vdec, corr, autocorr, z, zz] = th_moments(dr,options_,M_)
% Computes theoretical moments for GSA
%
% INPUTS
% - dr [structure] model information structure
% - options_ [structure] Matlab's structure describing the current options
% - M_ [structure] Matlab's structure describing the model
%
% OUTPUTS
% - vdec [double] variance decomposition matrix
% - corr [double] correlation matrix
% - autocorr [cell] contains autocorrelation or
% auto- and cross-covariance matrices
% - z [double] matrix containing mean, standard
% deviation, and variance vector
% - zz [double] autocorrelation matrix
% Copyright © 2012-2018 Dynare Team
% Copyright © 2012-2023 Dynare Team
%
% This file is part of Dynare.
%
@ -18,18 +33,16 @@ function [vdec, corr, autocorr, z, zz] = th_moments(dr,var_list)
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
global M_ oo_ options_
nvar = length(var_list);
nvar = length(options_.varobs);
if nvar == 0
nvar = length(dr.order_var);
ivar = [1:nvar]';
else
ivar=zeros(nvar,1);
for i=1:nvar
i_tmp = strmatch(var_list{i}, M_.endo_names, 'exact');
i_tmp = strmatch(options_.varobs{i}, M_.endo_names, 'exact');
if isempty(i_tmp)
error(['One of the variables specified does not exist']) ;
error('th_moments: One of the variables specified does not exist');
else
ivar(i) = i_tmp;
end
@ -39,21 +52,11 @@ end
[gamma_y,stationary_vars] = th_autocovariances(dr,ivar,M_, options_);
m = dr.ys(ivar(stationary_vars));
% i1 = find(abs(diag(gamma_y{1})) > 1e-12);
i1 = [1:length(ivar)];
i1 = 1:length(ivar);
s2 = diag(gamma_y{1});
sd = sqrt(s2);
z = [ m sd s2 ];
mean = m;
var = gamma_y{1};
%'THEORETICAL MOMENTS';
%'MEAN','STD. DEV.','VARIANCE');
z;
%'VARIANCE DECOMPOSITION (in percent)';
if M_.exo_nbr>1
@ -69,6 +72,7 @@ else
corr = gamma_y{1}(i1,i1);
end
if options_.ar > 0
zz=NaN(length(ivar),options_.ar);
%'COEFFICIENTS OF AUTOCORRELATION';
for i=1:options_.ar
if options_.opt_gsa.useautocorr

34
matlab/gsa/trank.m
View File

@ -1,34 +0,0 @@
function yr = trank(y)
% yr = trank(y);
% yr is the rank transformation of y
%
% Written by Marco Ratto
% Joint Research Centre, The European Commission,
% marco.ratto@ec.europa.eu
%
% Reference:
% M. Ratto, Global Sensitivity Analysis for Macroeconomic models, MIMEO, 2006.
% Copyright © 2012 European Commission
% Copyright © 2012-2017 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 <https://www.gnu.org/licenses/>.
[nr, nc] = size(y);
for j=1:nc
[dum, is]=sort(y(:,j));
yr(is,j)=[1:nr]'./nr;
end

18
matlab/ident_bruteforce.m
View File

@ -1,5 +1,5 @@
function [pars, cosnJ] = ident_bruteforce(J, max_dim_cova_group, TeX, name_tex, tittxt, tol_deriv)
% function [pars, cosnJ] = ident_bruteforce(J,n,TeX, pnames_TeX,tittxt)
function [pars, cosnJ] = ident_bruteforce(dname,fname,J, max_dim_cova_group, TeX, name_tex, tittxt, tol_deriv)
% [pars, cosnJ] = ident_bruteforce(dname,fname,J, max_dim_cova_group, TeX, name_tex, tittxt, tol_deriv)
% -------------------------------------------------------------------------
% given the Jacobian matrix J of moment derivatives w.r.t. parameters
% computes, for each column of J, the groups of columns from 1 to n that
@ -20,7 +20,7 @@ function [pars, cosnJ] = ident_bruteforce(J, max_dim_cova_group, TeX, name_tex,
% This function is called by
% * identification_analysis.m
% =========================================================================
% Copyright © 2009-2019 Dynare Team
% Copyright © 2009-2023 Dynare Team
%
% This file is part of Dynare.
%
@ -38,20 +38,18 @@ function [pars, cosnJ] = ident_bruteforce(J, max_dim_cova_group, TeX, name_tex,
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
% =========================================================================
global M_ options_
OutputDirectoryName = CheckPath('identification',M_.dname);
OutputDirectoryName = CheckPath('identification',dname);
totparam_nbr = size(J,2); % number of parameters
if nargin<2 || isempty(max_dim_cova_group)
if nargin<4 || isempty(max_dim_cova_group)
max_dim_cova_group = 4; % max n-tuple
end
if nargin<3 || isempty(TeX)
if nargin<5 || isempty(TeX)
TeX = 0; % no Tex output
tittxt='';
end
if nargin < 6
if nargin < 8
tol_deriv = 1.e-8;
end
@ -87,7 +85,7 @@ for ll = 1:max_dim_cova_group
end
dyn_waitbar_close(h);
if TeX
filename = [OutputDirectoryName '/' M_.fname '_collin_patterns_',tittxt1,'_' int2str(ll) '.tex'];
filename = [OutputDirectoryName '/' fname '_collin_patterns_',tittxt1,'_' int2str(ll) '.tex'];
fidTeX = fopen(filename,'w');
fprintf(fidTeX,'%% TeX-table generated by ident_bruteforce (Dynare).\n');
fprintf(fidTeX,['%% Collinearity patterns with ',int2str(ll),' parameter(s): ',tittxt,'\n']);

16
matlab/identification_analysis.m
View File

@ -1,5 +1,5 @@
function [ide_moments, ide_spectrum, ide_minimal, ide_hess, ide_reducedform, ide_dynamic, derivatives_info, info, error_indicator] = identification_analysis(params, indpmodel, indpstderr, indpcorr, options_ident, dataset_info, prior_exist, init)
% [ide_moments, ide_spectrum, ide_minimal, ide_hess, ide_reducedform, ide_dynamic, derivatives_info, info, error_indicator] = identification_analysis(params, indpmodel, indpstderr, indpcorr, options_ident, dataset_info, prior_exist, init)
function [ide_moments, ide_spectrum, ide_minimal, ide_hess, ide_reducedform, ide_dynamic, derivatives_info, info, error_indicator] = identification_analysis(M_,options_,oo_,bayestopt_,estim_params_,params, indpmodel, indpstderr, indpcorr, options_ident, dataset_info, prior_exist, init)
% [ide_moments, ide_spectrum, ide_minimal, ide_hess, ide_reducedform, ide_dynamic, derivatives_info, info, error_indicator] = identification_analysis(M_,options_,oo_,bayestopt_,estim_params_,params, indpmodel, indpstderr, indpcorr, options_ident, dataset_info, prior_exist, init)
% -------------------------------------------------------------------------
% This function wraps all identification analysis, i.e. it
% (1) wraps functions for the theoretical identification analysis based on moments (Iskrev, 2010),
@ -12,6 +12,11 @@ function [ide_moments, ide_spectrum, ide_minimal, ide_hess, ide_reducedform, ide
% moments, spectrum, reduced-form solution and dynamic model derivatives
% =========================================================================
% INPUTS
% * M_ [structure] describing the model
% * options_ [structure] describing the options
% * oo_ [structure] storing the results
% * bayestopt_ [structure] describing the priors
% * estim_params_ [structure] characterizing parameters to be estimated
% * params [mc_sample_nbr by totparam_nbr]
% parameter values for identification checks
% * indpmodel [modparam_nbr by 1]
@ -91,7 +96,6 @@ function [ide_moments, ide_spectrum, ide_minimal, ide_hess, ide_reducedform, ide
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
% =========================================================================
global oo_ M_ options_ bayestopt_ estim_params_
persistent ind_dMOMENTS ind_dREDUCEDFORM ind_dDYNAMIC
% persistent indices are necessary, because in a MC loop the numerical threshold
% used may provide vectors of different length, leading to crashes in MC loops
@ -138,7 +142,7 @@ error_indicator.identification_spectrum=0;
if info(1) == 0 %no errors in solution
% Compute parameter Jacobians for identification analysis
[MEAN, dMEAN, REDUCEDFORM, dREDUCEDFORM, DYNAMIC, dDYNAMIC, MOMENTS, dMOMENTS, dSPECTRUM, dSPECTRUM_NO_MEAN, dMINIMAL, derivatives_info] = get_identification_jacobians(estim_params_, M_, options_, options_ident, indpmodel, indpstderr, indpcorr, indvobs, oo_.dr, oo_.steady_state, oo_.exo_steady_state, oo_.exo_det_steady_state);
[~, ~, REDUCEDFORM, dREDUCEDFORM, DYNAMIC, dDYNAMIC, MOMENTS, dMOMENTS, dSPECTRUM, dSPECTRUM_NO_MEAN, dMINIMAL, derivatives_info] = get_identification_jacobians(estim_params_, M_, options_, options_ident, indpmodel, indpstderr, indpcorr, indvobs, oo_.dr, oo_.steady_state, oo_.exo_steady_state, oo_.exo_det_steady_state);
if isempty(dMINIMAL)
% Komunjer and Ng is not computed if (1) minimality conditions are not fullfilled or (2) there are more shocks and measurement errors than observables, so we need to reset options
error_indicator.identification_minimal = 1;
@ -296,7 +300,7 @@ if info(1) == 0 %no errors in solution
derivatives_info.no_DLIK = 1;
bounds = prior_bounds(bayestopt_, options_.prior_trunc); %reset bounds as lb and ub must only be operational during mode-finding
%note that for order>1 we do not provide any information on DT,DYss,DOM in derivatives_info, such that dsge_likelihood creates an error. Therefore the computation will be based on simulated_moment_uncertainty for order>1.
[fval, info, cost_flag, DLIK, AHess, ys, trend_coeff, M_, options_, bayestopt_, oo_.dr] = dsge_likelihood(params', dataset_, dataset_info, options_, M_, estim_params_, bayestopt_, bounds, oo_.dr, oo_.steady_state,oo_.exo_steady_state, oo_.exo_det_steady_state. derivatives_info); %non-used output variables need to be set for octave for some reason
[~, info, ~, ~, AHess, ~, ~, M_, options_, ~, oo_.dr] = dsge_likelihood(params', dataset_, dataset_info, options_, M_, estim_params_, bayestopt_, bounds, oo_.dr, oo_.steady_state,oo_.exo_steady_state, oo_.exo_det_steady_state. derivatives_info); %non-used output variables need to be set for octave for some reason
%note that for the order of parameters in AHess we have: stderr parameters come first, corr parameters second, model parameters third. the order within these blocks corresponds to the order specified in the estimated_params block
options_.analytic_derivation = analytic_derivation; %reset option
AHess = -AHess; %take negative of hessian
@ -462,7 +466,7 @@ if info(1) == 0 %no errors in solution
if advanced
% here we do not normalize (i.e. we set norm_dMOMENTS=1) as the OLS in ident_bruteforce is very sensitive to norm_dMOMENTS
[ide_moments.pars, ide_moments.cosndMOMENTS] = ident_bruteforce(dMOMENTS(ind_dMOMENTS,:), max_dim_cova_group, options_.TeX, options_ident.name_tex, options_ident.tittxt, tol_deriv);
[ide_moments.pars, ide_moments.cosndMOMENTS] = ident_bruteforce(M_.dname,M_.fname,dMOMENTS(ind_dMOMENTS,:), max_dim_cova_group, options_.TeX, options_ident.name_tex, options_ident.tittxt, tol_deriv);
end
%here we focus on the unnormalized S and V, which is then used in plot_identification.m and for prior_mc > 1

9
matlab/mh_autocorrelation_function.m
View File

@ -1,10 +1,9 @@
function mh_autocorrelation_function(options_,M_,estim_params_,type,blck,name1,name2)
% mh_autocorrelation_function(options_,M_,estim_params_,type,blck,name1,name2)
% This function plots the autocorrelation of the sampled draws in the
% posterior distribution.
%
%
% INPUTS
%
% options_ [structure] Dynare structure.
% M_ [structure] Dynare structure (related to model definition).
% estim_params_ [structure] Dynare structure (related to estimation).
@ -63,7 +62,7 @@ clear record;
PosteriorDraws = GetAllPosteriorDraws(M_.dname, M_.fname, column, FirstMhFile, FirstLine, TotalNumberOfMhFiles, NumberOfDraws, nblck, blck);
% Compute the autocorrelation function:
[autocov,autocor] = sample_autocovariance(PosteriorDraws,options_.mh_autocorrelation_function_size);
[~,autocor] = sample_autocovariance(PosteriorDraws,options_.mh_autocorrelation_function_size);
% Plot the posterior draws:
@ -101,9 +100,9 @@ axis tight
if ~exist(M_.dname, 'dir')
mkdir('.',M_.dname);
end
if ~exist([M_.dname filesep 'graphs'])
if ~exist([M_.dname filesep 'graphs'],'dir')
mkdir(M_.dname,'graphs');
end
plot_name=get_the_name(column,0,M_,estim_params_,options_);
plot_name=get_the_name(column,0,M_,estim_params_,options_.varobs);
dyn_saveas(hh_fig,[M_.dname, filesep, 'graphs', filesep, 'MH_Autocorrelation_' plot_name],options_.nodisplay,options_.graph_format)

2
matlab/mode_check.m
View File

@ -95,7 +95,7 @@ for plt = 1:nbplt
for k = 1:min(nstar,length(xparam)-(plt-1)*nstar)
subplot(nr,nc,k)
kk = (plt-1)*nstar+k;
[name,texname] = get_the_name(kk,options_.TeX,M_,estim_params_,options_);
[name,texname] = get_the_name(kk,options_.TeX,M_,estim_params_,options_.varobs);
xx = xparam;
if xparam(kk)~=0 && ~isinf(bounds.lb(kk)) && ~isinf(bounds.ub(kk))
l1 = max(bounds.lb(kk),(1-sign(xparam(kk))*ll)*xparam(kk)); m1 = 0; % lower bound

177
matlab/myboxplot.m
View File

@ -1,177 +0,0 @@
function sout = myboxplot (data,notched,symbol,vertical,maxwhisker)
% sout = myboxplot (data,notched,symbol,vertical,maxwhisker)
%
% Copyright © 2010-2017 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 <https://www.gnu.org/licenses/>.
% % % % endif
if nargin < 5 || isempty(maxwhisker), maxwhisker = 1.5; end
if nargin < 4 || isempty(vertical), vertical = 1; end
if nargin < 3 || isempty(symbol), symbol = ['+','o']; end
if nargin < 2 || isempty(notched), notched = 0; end
if length(symbol)==1, symbol(2)=symbol(1); end
if notched==1, notched=0.25; end
a=1-notched;
% ## figure out how many data sets we have
if iscell(data)
nc = length(data);
else
% if isvector(data), data = data(:); end
nc = size(data,2);
end
% ## compute statistics
% ## s will contain
% ## 1,5 min and max
% ## 2,3,4 1st, 2nd and 3rd quartile
% ## 6,7 lower and upper confidence intervals for median
s = zeros(7,nc);
box = zeros(1,nc);
whisker_x = ones(2,1)*[1:nc,1:nc];
whisker_y = zeros(2,2*nc);
outliers_x = [];
outliers_y = [];
outliers2_x = [];
outliers2_y = [];
for i=1:nc
% ## Get the next data set from the array or cell array
if iscell(data)
col = data{i}(:);
else
col = data(:,i);
end
% ## Skip missing data
% % % % % % % col(isnan(col) | isna (col)) = [];
col(isnan(col)) = [];
% ## Remember the data length
nd = length(col);
box(i) = nd;
if (nd > 1)
% ## min,max and quartiles
% s(1:5,i) = statistics(col)(1:5);
s(1,i)=min(col);
s(5,i)=max(col);
s(2,i)=myprctilecol(col,25);
s(3,i)=myprctilecol(col,50);
s(4,i)=myprctilecol(col,75);
% ## confidence interval for the median
est = 1.57*(s(4,i)-s(2,i))/sqrt(nd);
s(6,i) = max([s(3,i)-est, s(2,i)]);
s(7,i) = min([s(3,i)+est, s(4,i)]);
% ## whiskers out to the last point within the desired inter-quartile range
IQR = maxwhisker*(s(4,i)-s(2,i));
whisker_y(:,i) = [min(col(col >= s(2,i)-IQR)); s(2,i)];
whisker_y(:,nc+i) = [max(col(col <= s(4,i)+IQR)); s(4,i)];
% ## outliers beyond 1 and 2 inter-quartile ranges
outliers = col((col < s(2,i)-IQR & col >= s(2,i)-2*IQR) | (col > s(4,i)+IQR & col <= s(4,i)+2*IQR));
outliers2 = col(col < s(2,i)-2*IQR | col > s(4,i)+2*IQR);
outliers_x = [outliers_x; i*ones(size(outliers))];
outliers_y = [outliers_y; outliers];
outliers2_x = [outliers2_x; i*ones(size(outliers2))];
outliers2_y = [outliers2_y; outliers2];
elseif (nd == 1)
% ## all statistics collapse to the value of the point
s(:,i) = col;
% ## single point data sets are plotted as outliers.
outliers_x = [outliers_x; i];
outliers_y = [outliers_y; col];
else
% ## no statistics if no points
s(:,i) = NaN;
end
end
% % % % if isempty(outliers2_y)
% % % % outliers2_y=
% ## Note which boxes don't have enough stats
chop = find(box <= 1);
% ## Draw a box around the quartiles, with width proportional to the number of
% ## items in the box. Draw notches if desired.
box = box*0.23/max(box);
quartile_x = ones(11,1)*[1:nc] + [-a;-1;-1;1;1;a;1;1;-1;-1;-a]*box;
quartile_y = s([3,7,4,4,7,3,6,2,2,6,3],:);
% ## Draw a line through the median
median_x = ones(2,1)*[1:nc] + [-a;+a]*box;
% median_x=median(col);
median_y = s([3,3],:);
% ## Chop all boxes which don't have enough stats
quartile_x(:,chop) = [];
quartile_y(:,chop) = [];
whisker_x(:,[chop,chop+nc]) = [];
whisker_y(:,[chop,chop+nc]) = [];
median_x(:,chop) = [];
median_y(:,chop) = [];
% % % %
% ## Add caps to the remaining whiskers
cap_x = whisker_x;
cap_x(1,:) =cap_x(1,:)- 0.05;
cap_x(2,:) =cap_x(2,:)+ 0.05;
cap_y = whisker_y([1,1],:);
% #quartile_x,quartile_y
% #whisker_x,whisker_y
% #median_x,median_y
% #cap_x,cap_y
%
% ## Do the plot
mm=min(min(data));
MM=max(max(data));
if isnan(mm), mm=0; MM=0; end
if vertical
plot (quartile_x, quartile_y, 'b', ...
whisker_x, whisker_y, 'b--', ...
cap_x, cap_y, 'k', ...
median_x, median_y, 'r', ...
outliers_x, outliers_y, [symbol(1),'r'], ...
outliers2_x, outliers2_y, [symbol(2),'r']);
set(gca,'XTick',1:nc);
set(gca, 'XLim', [0.5, nc+0.5]);
set(gca, 'YLim', [mm-(MM-mm)*0.05-eps, MM+(MM-mm)*0.05+eps]);
else
% % % % % plot (quartile_y, quartile_x, "b;;",
% % % % % whisker_y, whisker_x, "b;;",
% % % % % cap_y, cap_x, "b;;",
% % % % % median_y, median_x, "r;;",
% % % % % outliers_y, outliers_x, [symbol(1),"r;;"],
% % % % % outliers2_y, outliers2_x, [symbol(2),"r;;"]);
end
if nargout
sout=s;
end
% % % endfunction

4
matlab/nonlinear-filters/online_auxiliary_filter.m
View File

@ -389,7 +389,7 @@ for plt = 1:nbplt
hh_fig = dyn_figure(options_.nodisplay,'Name','Parameters Trajectories');
for k=1:length(pmean)
subplot(nr,nc,k)
[name,texname] = get_the_name(k,TeX,M_,estim_params_,options_);
[name,texname] = get_the_name(k,TeX,M_,estim_params_,options_.varobs);
% Draw the surface for an interval containing 95% of the particles.
area(1:sample_size, ub95_xparam(k,:), 'FaceColor', [.9 .9 .9], 'BaseValue', min(lb95_xparam(k,:)));
hold on
@ -426,7 +426,7 @@ for plt = 1:nbplt
hh_fig = dyn_figure(options_.nodisplay,'Name','Parameters Densities');
for k=1:length(pmean)
subplot(nr,nc,k)
[name,texname] = get_the_name(k,TeX,M_,estim_params_,options_);
[name,texname] = get_the_name(k,TeX,M_,estim_params_,options_.varobs);
optimal_bandwidth = mh_optimal_bandwidth(xparam(k,:)',number_of_particles,bandwidth,kernel_function);
[density(:,1),density(:,2)] = kernel_density_estimate(xparam(k,:)', number_of_grid_points, ...
number_of_particles, optimal_bandwidth, kernel_function);

2
matlab/optimize_prior.m
View File

@ -65,7 +65,7 @@ disp('PRIOR OPTIMIZATION')
disp('------------------')
skipline()
for i = 1:length(xparams)
disp(['deep parameter ' int2str(i) ': ' get_the_name(i,0,M_,estim_params_,options_) '.'])
disp(['deep parameter ' int2str(i) ': ' get_the_name(i,0,M_,estim_params_,options_.varobs) '.'])
disp([' Initial condition ....... ' num2str(xinit(i)) '.'])
disp([' Prior mode .............. ' num2str(bayestopt_.p5(i)) '.'])
disp([' Optimized prior mode .... ' num2str(xparams(i)) '.'])

146
matlab/plot_identification.m
View File

@ -1,7 +1,8 @@
function plot_identification(params, idemoments, idehess, idemodel, idelre, advanced, tittxt, name, IdentifDirectoryName, tit_TeX, name_tex)
% function plot_identification(params,idemoments,idehess,idemodel, idelre, advanced, tittxt, name, IdentifDirectoryName)
function plot_identification(M_, params, idemoments, idehess, idemodel, idelre, advanced, tittxt, name, IdentifDirectoryName, fname, options_, estim_params_, bayestopt_, tit_TeX, name_tex)
% plot_identification(M_, params,idemoments,idehess,idemodel, idelre, advanced, tittxt, name, IdentifDirectoryName, fname, options_, estim_params_, bayestopt_, tit_TeX, name_tex)
%
% INPUTS
% o M_ [structure] model
% o params [array] parameter values for identification checks
% o idemoments [structure] identification results for the moments
% o idehess [structure] identification results for the Hessian
@ -9,10 +10,15 @@ function plot_identification(params, idemoments, idehess, idemodel, idelre, adva
% o idelre [structure] identification results for the LRE model
% o advanced [integer] flag for advanced identification checks
% o tittxt [char] name of the results to plot
% o name [char] list of names
% o name [char] list of parameter names
% o IdentifDirectoryName [char] directory name
% o fname [char] file name
% o options_ [structure] structure describing the current options
% o estim_params_ [structure] characterizing parameters to be estimated
% o bayestopt_ [structure] describing the priors
% o tittxt [char] TeX-name of the results to plot
% o name_tex [char] TeX-names of the parameters
%
% OUTPUTS
% None
%
@ -36,13 +42,11 @@ function plot_identification(params, idemoments, idehess, idemodel, idelre, adva
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
global M_ options_
if nargin <10 || isempty(tit_TeX)
if nargin <14 || isempty(tit_TeX)
tit_TeX=tittxt;
end
if nargin <11
if nargin <15
name_tex=name;
end
@ -54,20 +58,19 @@ si_dLREnorm = idelre.si_dDYNAMICnorm;
tittxt1=regexprep(tittxt, ' ', '_');
tittxt1=strrep(tittxt1, '.', '');
if SampleSize == 1
si_dMOMENTS = idemoments.si_dMOMENTS;
hh_fig = dyn_figure(options_.nodisplay,'Name',[tittxt, ' - Identification using info from observables']);
subplot(211)
mmm = (idehess.ide_strength_dMOMENTS);
[ss, is] = sort(mmm);
[~, is] = sort(mmm);
if ~all(isnan(idehess.ide_strength_dMOMENTS_prior)) ...
&& ~(nparam == 1 && ~isoctave && matlab_ver_less_than('9.7')) % MATLAB < R2019b does not accept bar(1, [2 3])
bar(1:nparam,log([idehess.ide_strength_dMOMENTS(:,is)' idehess.ide_strength_dMOMENTS_prior(:,is)']))
else
bar(1:nparam,log([idehess.ide_strength_dMOMENTS(:,is)' ]))
bar(1:nparam,log(idehess.ide_strength_dMOMENTS(:,is)'))
end
hold on
plot((1:length(idehess.ide_strength_dMOMENTS(:,is)))-0.15,log([idehess.ide_strength_dMOMENTS(:,is)']),'o','MarkerSize',7,'MarkerFaceColor',[0 0 0],'MarkerEdgeColor','none')
plot((1:length(idehess.ide_strength_dMOMENTS_prior(:,is)))+0.15,log([idehess.ide_strength_dMOMENTS_prior(:,is)']),'o','MarkerSize',7,'MarkerFaceColor',[0 0 0],'MarkerEdgeColor','none')
plot((1:length(idehess.ide_strength_dMOMENTS(:,is)))-0.15,log(idehess.ide_strength_dMOMENTS(:,is)'),'o','MarkerSize',7,'MarkerFaceColor',[0 0 0],'MarkerEdgeColor','none')
plot((1:length(idehess.ide_strength_dMOMENTS_prior(:,is)))+0.15,log(idehess.ide_strength_dMOMENTS_prior(:,is)'),'o','MarkerSize',7,'MarkerFaceColor',[0 0 0],'MarkerEdgeColor','none')
if any(isinf(log(idehess.ide_strength_dMOMENTS(idehess.identified_parameter_indices))))
%-Inf, i.e. 0 strength
inf_indices=find(isinf(log(idehess.ide_strength_dMOMENTS(idehess.identified_parameter_indices))) & log(idehess.ide_strength_dMOMENTS(idehess.identified_parameter_indices))<0);
@ -117,8 +120,8 @@ if SampleSize == 1
bar(1:nparam, log([idehess.deltaM(is)]))
end
hold on
plot((1:length(idehess.deltaM(is)))-0.15,log([idehess.deltaM(is)']),'o','MarkerSize',7,'MarkerFaceColor',[0 0 0],'MarkerEdgeColor','none')
plot((1:length(idehess.deltaM_prior(is)))+0.15,log([idehess.deltaM_prior(is)']),'o','MarkerSize',7,'MarkerFaceColor',[0 0 0],'MarkerEdgeColor','none')
plot((1:length(idehess.deltaM(is)))-0.15,log(idehess.deltaM(is)'),'o','MarkerSize',7,'MarkerFaceColor',[0 0 0],'MarkerEdgeColor','none')
plot((1:length(idehess.deltaM_prior(is)))+0.15,log(idehess.deltaM_prior(is)'),'o','MarkerSize',7,'MarkerFaceColor',[0 0 0],'MarkerEdgeColor','none')
inf_pos=find(isinf(log(idehess.deltaM)));
if ~isempty(inf_pos)
inf_indices=~ismember(inf_pos,idehess.sensitivity_zero_pos);
@ -152,19 +155,19 @@ if SampleSize == 1
title('Sensitivity component with moments Information matrix (log-scale)')
end
if options_.TeX && any(strcmp('eps',cellstr(options_.graph_format)))
fidTeX = fopen([IdentifDirectoryName '/' M_.fname '_ident_strength_' tittxt1,'.tex'],'w');
fidTeX = fopen([IdentifDirectoryName '/' fname '_ident_strength_' tittxt1,'.tex'],'w');
fprintf(fidTeX,'%% TeX eps-loader file generated by plot_identification.m (Dynare).\n');
fprintf(fidTeX,['%% ' datestr(now,0) '\n\n']);
fprintf(fidTeX,'\\begin{figure}[H]\n');
fprintf(fidTeX,'\\centering \n');
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_ident_strength_%s}\n',[IdentifDirectoryName '/' M_.fname],tittxt1);
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_ident_strength_%s}\n',[IdentifDirectoryName '/' fname],tittxt1);
fprintf(fidTeX,'\\caption{%s - Identification using info from observables.}',tit_TeX);
fprintf(fidTeX,'\\label{Fig:ident:%s}\n',deblank(tittxt));
fprintf(fidTeX,'\\end{figure}\n\n');
fprintf(fidTeX,'%% End Of TeX file. \n');
fclose(fidTeX);
end
dyn_saveas(hh_fig,[IdentifDirectoryName '/' M_.fname '_ident_strength_' tittxt1],options_.nodisplay,options_.graph_format);
dyn_saveas(hh_fig,[IdentifDirectoryName '/' fname '_ident_strength_' tittxt1],options_.nodisplay,options_.graph_format);
if advanced
if ~options_.nodisplay
@ -197,14 +200,14 @@ if SampleSize == 1
end
legend('Moments','Model','LRE model','Location','Best')
title('Sensitivity bars using derivatives (log-scale)')
dyn_saveas(hh_fig,[IdentifDirectoryName '/' M_.fname '_sensitivity_' tittxt1 ],options_.nodisplay,options_.graph_format);
dyn_saveas(hh_fig,[IdentifDirectoryName '/' fname '_sensitivity_' tittxt1 ],options_.nodisplay,options_.graph_format);
if options_.TeX && any(strcmp('eps',cellstr(options_.graph_format)))
fidTeX = fopen([IdentifDirectoryName '/' M_.fname '_sensitivity_' tittxt1,'.tex'],'w');
fidTeX = fopen([IdentifDirectoryName '/' fname '_sensitivity_' tittxt1,'.tex'],'w');
fprintf(fidTeX,'%% TeX eps-loader file generated by plot_identification.m (Dynare).\n');
fprintf(fidTeX,['%% ' datestr(now,0) '\n\n']);
fprintf(fidTeX,'\\begin{figure}[H]\n');
fprintf(fidTeX,'\\centering \n');
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_sensitivity_%s}\n',[IdentifDirectoryName '/' M_.fname],tittxt1);
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_sensitivity_%s}\n',[IdentifDirectoryName '/' fname],tittxt1);
fprintf(fidTeX,'\\caption{%s - Sensitivity plot.}',tit_TeX);
fprintf(fidTeX,'\\label{Fig:sensitivity:%s}\n',deblank(tittxt));
fprintf(fidTeX,'\\end{figure}\n\n');
@ -215,9 +218,6 @@ if SampleSize == 1
% identificaton patterns
for j=1:size(idemoments.cosndMOMENTS,2)
pax=NaN(nparam,nparam);
% fprintf('\n')
% disp(['Collinearity patterns with ', int2str(j) ,' parameter(s)'])
% fprintf('%-15s [%-*s] %10s\n','Parameter',(15+1)*j,' Expl. params ','cosn')
for i=1:nparam
namx='';
for in=1:j
@ -233,7 +233,6 @@ if SampleSize == 1
pax(i,dumpindx)=idemoments.cosndMOMENTS(i,j);
end
end
% fprintf('%-15s [%s] %10.3f\n',name{i},namx,idemoments.cosndMOMENTS(i,j))
end
hh_fig = dyn_figure(options_.nodisplay,'Name',[tittxt,' - Collinearity patterns with ', int2str(j) ,' parameter(s)']);
imagesc(pax,[0 1]);
@ -260,14 +259,14 @@ if SampleSize == 1
set(gca,'xgrid','on')
set(gca,'ygrid','on')
xlabel([tittxt,' - Collinearity patterns with ', int2str(j) ,' parameter(s)'],'interpreter','none')
dyn_saveas(hh_fig,[ IdentifDirectoryName '/' M_.fname '_ident_collinearity_' tittxt1 '_' int2str(j) ],options_.nodisplay,options_.graph_format);
dyn_saveas(hh_fig,[ IdentifDirectoryName '/' fname '_ident_collinearity_' tittxt1 '_' int2str(j) ],options_.nodisplay,options_.graph_format);
if options_.TeX && any(strcmp('eps',cellstr(options_.graph_format)))
fidTeX = fopen([ IdentifDirectoryName '/' M_.fname '_ident_collinearity_' tittxt1 '_' int2str(j),'.tex'],'w');
fidTeX = fopen([ IdentifDirectoryName '/' fname '_ident_collinearity_' tittxt1 '_' int2str(j),'.tex'],'w');
fprintf(fidTeX,'%% TeX eps-loader file generated by plot_identification.m (Dynare).\n');
fprintf(fidTeX,['%% ' datestr(now,0) '\n\n']);
fprintf(fidTeX,'\\begin{figure}[H]\n');
fprintf(fidTeX,'\\centering \n');
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_ident_collinearity_%s_%u}\n',[IdentifDirectoryName '/' M_.fname],tittxt1,j);
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_ident_collinearity_%s_%u}\n',[IdentifDirectoryName '/' fname],tittxt1,j);
fprintf(fidTeX,'\\caption{%s - Collinearity patterns with %u parameter(s).}',tit_TeX,j);
fprintf(fidTeX,'\\label{Fig:collinearity:%s:%u_pars}\n',deblank(tittxt),j);
fprintf(fidTeX,'\\end{figure}\n\n');
@ -276,7 +275,7 @@ if SampleSize == 1
end
end
skipline()
[U,S,V]=svd(idehess.AHess,0);
[~,S,V]=svd(idehess.AHess,0);
S=diag(S);
if idehess.flag_score
if nparam<5
@ -289,8 +288,6 @@ if SampleSize == 1
tex_tit_2=[tittxt,' - Identification patterns (Information matrix): HIGHEST SV'];
end
else
% S = idemoments.S;
% V = idemoments.V;
if nparam<5
f1 = dyn_figure(options_.nodisplay,'Name',[tittxt,' - Identification patterns (moments Information matrix)']);
tex_tit_1=[tittxt,' - Identification patterns (moments Information matrix)'];
@ -329,14 +326,14 @@ if SampleSize == 1
end
title(['Singular value ',num2str(Stit)])
end
dyn_saveas(f1,[ IdentifDirectoryName '/' M_.fname '_ident_pattern_' tittxt1 '_1' ],options_.nodisplay,options_.graph_format);
dyn_saveas(f1,[ IdentifDirectoryName '/' fname '_ident_pattern_' tittxt1 '_1' ],options_.nodisplay,options_.graph_format);
if options_.TeX && any(strcmp('eps',cellstr(options_.graph_format)))
fidTeX = fopen([ IdentifDirectoryName '/' M_.fname '_ident_pattern_' tittxt1 '_1','.tex'],'w');
fidTeX = fopen([ IdentifDirectoryName '/' fname '_ident_pattern_' tittxt1 '_1','.tex'],'w');
fprintf(fidTeX,'%% TeX eps-loader file generated by plot_identification.m (Dynare).\n');
fprintf(fidTeX,['%% ' datestr(now,0) '\n\n']);
fprintf(fidTeX,'\\begin{figure}[H]\n');
fprintf(fidTeX,'\\centering \n');
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_ident_pattern_%s_1}\n',[IdentifDirectoryName '/' M_.fname],tittxt1);
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_ident_pattern_%s_1}\n',[IdentifDirectoryName '/' fname],tittxt1);
fprintf(fidTeX,'\\caption{%s.}',tex_tit_1);
fprintf(fidTeX,'\\label{Fig:ident_pattern:%s:1}\n',tittxt1);
fprintf(fidTeX,'\\end{figure}\n\n');
@ -344,14 +341,14 @@ if SampleSize == 1
fclose(fidTeX);
end
if nparam>4
dyn_saveas(f2,[ IdentifDirectoryName '/' M_.fname '_ident_pattern_' tittxt1 '_2' ],options_.nodisplay,options_.graph_format);
dyn_saveas(f2,[ IdentifDirectoryName '/' fname '_ident_pattern_' tittxt1 '_2' ],options_.nodisplay,options_.graph_format);
if options_.TeX && any(strcmp('eps',cellstr(options_.graph_format)))
fidTeX = fopen([ IdentifDirectoryName '/' M_.fname '_ident_pattern_' tittxt1 '_2.tex'],'w');
fidTeX = fopen([ IdentifDirectoryName '/' fname '_ident_pattern_' tittxt1 '_2.tex'],'w');
fprintf(fidTeX,'%% TeX eps-loader file generated by plot_identification.m (Dynare).\n');
fprintf(fidTeX,['%% ' datestr(now,0) '\n\n']);
fprintf(fidTeX,'\\begin{figure}[H]\n');
fprintf(fidTeX,'\\centering \n');
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_ident_pattern_%s_2}\n',[IdentifDirectoryName '/' M_.fname],tittxt1);
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_ident_pattern_%s_2}\n',[IdentifDirectoryName '/' fname],tittxt1);
fprintf(fidTeX,'\\caption{%s.}',tex_tit_2);
fprintf(fidTeX,'\\label{Fig:ident_pattern:%s:2}\n',tittxt1);
fprintf(fidTeX,'\\end{figure}\n\n');
@ -362,10 +359,10 @@ if SampleSize == 1
end
else
hh_fig = dyn_figure(options_.nodisplay,'Name',['MC sensitivities']);
hh_fig = dyn_figure(options_.nodisplay,'Name','MC sensitivities');
subplot(211)
mmm = (idehess.ide_strength_dMOMENTS);
[ss, is] = sort(mmm);
[~, is] = sort(mmm);
mmm = mean(si_dMOMENTSnorm)';
mmm = mmm./max(mmm);
if advanced
@ -392,14 +389,14 @@ else
legend('Moments','Model','LRE model','Location','Best')
end
title('MC mean of sensitivity measures')
dyn_saveas(hh_fig,[ IdentifDirectoryName '/' M_.fname '_MC_sensitivity' ],options_.nodisplay,options_.graph_format);
dyn_saveas(hh_fig,[ IdentifDirectoryName '/' fname '_MC_sensitivity' ],options_.nodisplay,options_.graph_format);
if options_.TeX && any(strcmp('eps',cellstr(options_.graph_format)))
fidTeX = fopen([ IdentifDirectoryName '/' M_.fname '_MC_sensitivity.tex'],'w');
fidTeX = fopen([ IdentifDirectoryName '/' fname '_MC_sensitivity.tex'],'w');
fprintf(fidTeX,'%% TeX eps-loader file generated by plot_identification.m (Dynare).\n');
fprintf(fidTeX,['%% ' datestr(now,0) '\n\n']);
fprintf(fidTeX,'\\begin{figure}[H]\n');
fprintf(fidTeX,'\\centering \n');
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_MC_sensitivity}\n',[IdentifDirectoryName '/' M_.fname]);
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_MC_sensitivity}\n',[IdentifDirectoryName '/' fname]);
fprintf(fidTeX,'\\caption{MC mean of sensitivity measures}');
fprintf(fidTeX,'\\label{Fig:_MC_sensitivity}\n');
fprintf(fidTeX,'\\end{figure}\n\n');
@ -408,57 +405,62 @@ else
end
if advanced
if ~options_.nodisplay,
if ~options_.nodisplay
skipline()
disp('Displaying advanced diagnostics')
end
% options_.nograph=1;
hh_fig = dyn_figure(options_.nodisplay,'Name','MC Condition Number');
subplot(221)
hist(log10(idemodel.cond))
if isoctave
hist(log10(idemodel.cond))
else
histogram(log10(idemodel.cond))
end
title('log10 of Condition number in the model')
subplot(222)
hist(log10(idemoments.cond))
if isoctave
hist(log10(idemoments.cond))
else
histogram(log10(idemoments.cond))
end
title('log10 of Condition number in the moments')
subplot(223)
hist(log10(idelre.cond))
if isoctave
hist(log10(idelre.cond))
else
histogram(log10(idelre.cond))
end
title('log10 of Condition number in the LRE model')
dyn_saveas(hh_fig,[IdentifDirectoryName '/' M_.fname '_ident_COND' ],options_.nodisplay,options_.graph_format);
dyn_saveas(hh_fig,[IdentifDirectoryName '/' fname '_ident_COND' ],options_.nodisplay,options_.graph_format);
options_mcf.pvalue_ks = 0.1;
options_mcf.pvalue_corr = 0.001;
options_mcf.alpha2 = 0;
options_mcf.param_names = name;
options_mcf.param_names_tex = name_tex;
options_mcf.fname_ = M_.fname;
options_mcf.fname_ = fname;
options_mcf.OutputDirectoryName = IdentifDirectoryName;
options_mcf.beha_title = 'LOW condition nbr';
options_mcf.nobeha_title = 'HIGH condition nbr';
if options_.TeX
options_mcf.beha_title_latex = 'LOW condition nbr';
options_mcf.nobeha_title_latex = 'HIGH condition nbr';
end
options_mcf.amcf_name = 'MC_HighestCondNumberLRE';
options_mcf.amcf_title = 'MC Highest Condition Number LRE Model';
options_mcf.title = 'MC Highest Condition Number LRE Model';
ncut=floor(SampleSize/10*9);
[dum,is]=sort(idelre.cond);
mcf_analysis(params, is(1:ncut), is(ncut+1:end), options_mcf, options_);
[~,is]=sort(idelre.cond);
mcf_analysis(params, is(1:ncut), is(ncut+1:end), options_mcf, M_, options_, bayestopt_, estim_params_);
options_mcf.amcf_name = 'MC_HighestCondNumberModel';
options_mcf.amcf_title = 'MC Highest Condition Number Model Solution';
options_mcf.title = 'MC Highest Condition Number Model Solution';
[dum,is]=sort(idemodel.cond);
mcf_analysis(params, is(1:ncut), is(ncut+1:end), options_mcf, options_);
[~,is]=sort(idemodel.cond);
mcf_analysis(params, is(1:ncut), is(ncut+1:end), options_mcf, M_, options_, bayestopt_, estim_params_);
options_mcf.amcf_name = 'MC_HighestCondNumberMoments';
options_mcf.amcf_title = 'MC Highest Condition Number Model Moments';
options_mcf.title = 'MC Highest Condition Number Model Moments';
[dum,is]=sort(idemoments.cond);
mcf_analysis(params, is(1:ncut), is(ncut+1:end), options_mcf, options_);
% [proba, dproba] = stab_map_1(idemoments.Mco', is(1:ncut), is(ncut+1:end), 'HighestCondNumberMoments_vs_Mco', 1, [], IdentifDirectoryName);
% for j=1:nparam,
% % ibeh=find(idemoments.Mco(j,:)<0.9);
% % inonbeh=find(idemoments.Mco(j,:)>=0.9);
% % if ~isempty(ibeh) && ~isempty(inonbeh)
% % [proba, dproba] = stab_map_1(params, ibeh, inonbeh, ['HighestMultiCollinearity_',name{j}], 1, [], IdentifDirectoryName);
% % end
% [~,is]=sort(idemoments.Mco(:,j));
% [proba, dproba] = stab_map_1(params, is(1:ncut), is(ncut+1:end), ['MC_HighestMultiCollinearity_',name{j}], 1, [], IdentifDirectoryName, 0.15);
% end
[~,is]=sort(idemoments.cond);
mcf_analysis(params, is(1:ncut), is(ncut+1:end), options_mcf, M_, options_, bayestopt_, estim_params_);
if nparam<5
f1 = dyn_figure(options_.nodisplay,'Name',[tittxt,' - MC Identification patterns (moments): HIGHEST SV']);
@ -488,8 +490,10 @@ else
SSS = idemoments.S(:,jj);
end
subplot(nsubplo,1,jj)
post_median=NaN(1,nparam);
hpd_interval=NaN(nparam,2);
for i=1:nparam
[post_mean, post_median(:,i), post_var, hpd_interval(i,:), post_deciles] = posterior_moments(VVV(:,i),0,0.9);
[~, post_median(:,i), ~, hpd_interval(i,:)] = posterior_moments(VVV(:,i),0,0.9);
end
bar(post_median)
hold on, plot(hpd_interval,'--*r'),
@ -507,14 +511,14 @@ else
end
title(['MEAN Singular value ',num2str(Stit)])
end
dyn_saveas(f1,[IdentifDirectoryName '/' M_.fname '_MC_ident_pattern_1' ],options_.nodisplay,options_.graph_format);
dyn_saveas(f1,[IdentifDirectoryName '/' fname '_MC_ident_pattern_1' ],options_.nodisplay,options_.graph_format);
if options_.TeX && any(strcmp('eps',cellstr(options_.graph_format)))
fidTeX = fopen([IdentifDirectoryName '/' M_.fname '_MC_ident_pattern_1.tex'],'w');
fidTeX = fopen([IdentifDirectoryName '/' fname '_MC_ident_pattern_1.tex'],'w');
fprintf(fidTeX,'%% TeX eps-loader file generated by plot_identification.m (Dynare).\n');
fprintf(fidTeX,['%% ' datestr(now,0) '\n\n']);
fprintf(fidTeX,'\\begin{figure}[H]\n');
fprintf(fidTeX,'\\centering \n');
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_MC_ident_pattern_1}\n',[IdentifDirectoryName '/' M_.fname]);
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_MC_ident_pattern_1}\n',[IdentifDirectoryName '/' fname]);
fprintf(fidTeX,'\\caption{%s.}',tex_tit_1);
fprintf(fidTeX,'\\label{Fig:MC_ident_pattern:1}\n');
fprintf(fidTeX,'\\end{figure}\n\n');
@ -522,14 +526,14 @@ else
fclose(fidTeX);
end
if nparam>4
dyn_saveas(f2,[ IdentifDirectoryName '/' M_.fname '_MC_ident_pattern_2' ],options_.nodisplay,options_.graph_format);
dyn_saveas(f2,[ IdentifDirectoryName '/' fname '_MC_ident_pattern_2' ],options_.nodisplay,options_.graph_format);
if options_.TeX && any(strcmp('eps',cellstr(options_.graph_format)))
fidTeX = fopen([ IdentifDirectoryName '/' M_.fname '_MC_ident_pattern_2.tex'],'w');
fidTeX = fopen([ IdentifDirectoryName '/' fname '_MC_ident_pattern_2.tex'],'w');
fprintf(fidTeX,'%% TeX eps-loader file generated by plot_identification.m (Dynare).\n');
fprintf(fidTeX,['%% ' datestr(now,0) '\n\n']);
fprintf(fidTeX,'\\begin{figure}[H]\n');
fprintf(fidTeX,'\\centering \n');
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_MC_ident_pattern_2}\n',[IdentifDirectoryName '/' M_.fname]);
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_MC_ident_pattern_2}\n',[IdentifDirectoryName '/' fname]);
fprintf(fidTeX,'\\caption{%s.}',tex_tit_2);
fprintf(fidTeX,'\\label{Fig:MC_ident_pattern:2}\n');
fprintf(fidTeX,'\\end{figure}\n\n');

8
matlab/plot_priors.m
View File

@ -1,5 +1,5 @@
function plot_priors(bayestopt_,M_,estim_params_,options_,optional_title)
% function plot_priors
% plot_priors(bayestopt_,M_,estim_params_,options_,optional_title)
% plots prior density
%
% INPUTS
@ -41,7 +41,7 @@ else
figurename = optional_title;
end
npar = length(bayestopt_.p1);
[nbplt,nr,nc,lr,lc,nstar] = pltorg(npar);
[nbplt,nr,nc,~,~,nstar] = pltorg(npar);
if TeX && any(strcmp('eps',cellstr(options_.graph_format)))
fidTeX = fopen([latexDirectoryName filesep M_.fname '_Priors.tex'],'w');
@ -59,8 +59,8 @@ for plt = 1:nbplt
for index=1:nstar0
names = [];
i = (plt-1)*nstar + index;
[x,f,abscissa,dens,binf,bsup] = draw_prior_density(i,bayestopt_);
[nam,texnam] = get_the_name(i,TeX,M_,estim_params_,options_);
[x,f] = draw_prior_density(i,bayestopt_);
[nam,texnam] = get_the_name(i,TeX,M_,estim_params_,options_.varobs);
subplot(nr,nc,index)
hh_plt = plot(x,f,'-k','linewidth',2);
set(hh_plt,'color',[0.7 0.7 0.7]);

2
matlab/print_table_prior.m
View File

@ -52,7 +52,7 @@ options_.prior_trunc = prior_trunc_backup ;
RESIZE = false;
for i=1:size(bayestopt_.name,1)
[Name,tmp] = get_the_name(i,1,M_,estim_params_,options_);
[Name,~] = get_the_name(i,1,M_,estim_params_,options_.varobs);
if length(Name)>size(T1,2)
resize = true;
else

2
matlab/prior_draw.m
View File

@ -23,8 +23,6 @@ function pdraw = prior_draw(bayestopt_, prior_trunc, uniform)
%
% NOTE 1. Input arguments 1 and 2 are only needed for initialization.
% NOTE 2. A given draw from the joint prior distribution does not satisfy BK conditions a priori.
% NOTE 3. This code relies on bayestopt_ as created in the base workspace
% by the preprocessor (or as updated in subsequent pieces of code and handed to the base workspace)
%
% Copyright © 2006-2023 Dynare Team
%

8
matlab/set_prior.m
View File

@ -133,7 +133,7 @@ if ncn
bayestopt_.p4 = [ bayestopt_.p4; estim_params_.corrn(:,10)]; %take generalized distribution into account
bayestopt_.jscale = [ bayestopt_.jscale; estim_params_.corrn(:,11)];
baseid = length(bayestopt_.name);
bayestopt_.name = [bayestopt_.name; cell(ncn, 1)];;
bayestopt_.name = [bayestopt_.name; cell(ncn, 1)];
for i=1:ncn
k1 = estim_params_.corrn(i,1);
k2 = estim_params_.corrn(i,2);
@ -163,8 +163,8 @@ bayestopt_.p7 = bayestopt_.p6 ;
%% check for point priors and disallow them as they do not work with MCMC
if any(bayestopt_.p2 ==0)
error(sprintf(['Error in prior for %s: you cannot use a point prior in estimation. Either increase the prior standard deviation',...
' or fix the parameter completely.'], bayestopt_.name{bayestopt_.p2 ==0}))
error('Error in prior for %s: you cannot use a point prior in estimation. Either increase the prior standard deviation',...
' or fix the parameter completely.', bayestopt_.name{bayestopt_.p2 ==0})
end
% generalized location parameters by default for beta distribution
@ -285,7 +285,7 @@ end
CheckPath('prior',M_.dname);
% I save the prior definition if the prior has changed.
if exist([ M_.dname '/prior/definition.mat'])
if exist([ M_.dname '/prior/definition.mat'],'file')
old = load([M_.dname '/prior/definition.mat'],'bayestopt_');
prior_has_changed = 0;
if length(bayestopt_.p1)==length(old.bayestopt_.p1)

2
matlab/trace_plot.m
View File

@ -153,7 +153,7 @@ end
if strcmpi(type,'PosteriorDensity')
plot_name='Posterior';
else
plot_name=get_the_name(column,0,M_,estim_params_,options_);
plot_name=get_the_name(column,0,M_,estim_params_,options_.varobs);
end
if n_nblocks_to_plot==1
plot_name=[plot_name,'_blck_',num2str(blck)];

2
matlab/utilities/estimation/check_hessian_at_the_mode.m
View File

@ -48,7 +48,7 @@ catch
params_at_bound = find(abs(xparam1-bounds.ub)<tol_bounds | abs(xparam1-bounds.lb)<tol_bounds);
if ~isempty(params_at_bound)
for ii=1:length(params_at_bound)
params_at_bound_name{ii,1}=get_the_name(params_at_bound(ii),0,M_,estim_params_,options_);
params_at_bound_name{ii,1}=get_the_name(params_at_bound(ii),0,M_,estim_params_,options_.varobs);
end
disp_string=[params_at_bound_name{1,:}];
for ii=2:size(params_at_bound_name,1)

2
matlab/utilities/estimation/set_mcmc_prior_bounds.m
View File

@ -37,7 +37,7 @@ bounds = prior_bounds(bayestopt_, options_.prior_trunc);
outside_bound_pars = find(xparam < bounds.lb | xparam > bounds.ub);
if ~isempty(outside_bound_pars)
for ii = 1:length(outside_bound_pars)
outside_bound_par_names{ii,1} = get_the_name(ii,0,M_,estim_params_,options_);
outside_bound_par_names{ii,1} = get_the_name(ii,0,M_,estim_params_,options_.varobs);
end
disp_string = [outside_bound_par_names{1,:}];
for ii = 2:size(outside_bound_par_names,1)

8
matlab/write_latex_prior_table.m
View File

@ -28,7 +28,7 @@ function write_latex_prior_table
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
global M_ options_ bayestopt_ estim_params_
global M_ options_ estim_params_
if ~isbayes(estim_params_)
fprintf('\nwrite_latex_prior_table:: No prior distributions detected. Skipping table creation.\n')
@ -39,13 +39,13 @@ if (size(estim_params_.var_endo,1) || size(estim_params_.corrn,1))
% Prior over measurement errors are defined...
if ((isfield(options_,'varobs') && isempty(options_.varobs)) || ~isfield(options_,'varobs'))
% ... But the list of observed variabled is not yet defined.
fprintf(['\nwrite_latex_prior_table:: varobs should be declared before. Skipping table creation.\n'])
fprintf('\nwrite_latex_prior_table:: varobs should be declared before. Skipping table creation.\n')
return
end
end
% Fill or update bayestopt_ structure
[xparam1, estim_params_, BayesOptions, lb, ub, M_] = set_prior(estim_params_, M_, options_);
[~, estim_params_, BayesOptions, ~, ~, M_] = set_prior(estim_params_, M_, options_);
% Get untruncated bounds
bounds = prior_bounds(BayesOptions, options_.prior_trunc);
@ -112,7 +112,7 @@ fprintf(fidTeX,'\\endlastfoot\n');
% Column 8: the upper bound of the interval containing 90% of the prior mass.
PriorIntervals = prior_bounds(BayesOptions,(1-options_.prior_interval)/2) ;
for i=1:size(BayesOptions.name,1)
[tmp,TexName] = get_the_name(i, 1, M_, estim_params_, options_);
[~,TexName] = get_the_name(i, 1, M_, estim_params_, options_.varobs);
PriorShape = PriorNames{ BayesOptions.pshape(i) };
PriorMean = BayesOptions.p1(i);
PriorMode = BayesOptions.p5(i);

10
meson.build
View File

@ -804,10 +804,12 @@ mod_and_m_tests = [
'estimation/fsdat_simul.m' ] },
{ 'test' : [ 'estimation/fs2000.mod' ],
'extra' : [ 'estimation/fsdat_simul.m' ] },
{ 'test' : [ 'gsa/ls2003a.mod' ],
'extra' : [ 'gsa/data_ca1.m' ] },
{ 'test' : [ 'gsa/ls2003.mod' ],
'extra' : [ 'gsa/data_ca1.m' ] },
{ 'test' : [ 'gsa/ls2003a.mod',
'gsa/ls2003.mod',
'gsa/ls2003scr.mod',
'gsa/ls2003ide.mod' ],
'extra' : [ 'gsa/data_ca1.m',
'gsa/ls2003_model.inc' ] },
{ 'test' : [ 'gsa/cod_ML_morris/cod_ML_morris.mod' ] },
{ 'test' : [ 'gsa/morris/nk_est.mod' ],
'extra' : [ 'gsa/morris/nk_est_data.m' ] },

2
preprocessor

@ -1 +1 @@
Subproject commit 73be95423d3f8c1aa05a5edae2712ffbf646602a
Subproject commit 378d00fc3a50a11c3dce615ea337ad55989c938e

4
tests/TeX/fs2000_corr_ME.mod
View File

@ -185,6 +185,8 @@ stoch_simul(order=1,irf=20,graph_format=eps,periods=0,contemporaneous_correlatio
collect_latex_files;
//identification(advanced=1,max_dim_cova_group=3,prior_mc=250);
if system(['pdflatex -halt-on-error -interaction=batchmode ' M_.fname '_TeX_binder.tex'])
[status, cmdout]=system(['pdflatex -halt-on-error -interaction=nonstopmode ' M_.fname '_TeX_binder.tex']);
if status
cmdout
error('TeX-File did not compile.')
end

2
tests/analytic_derivatives/BrockMirman_PertParamsDerivs.mod
View File

@ -418,7 +418,7 @@ for jj = 1:2
if jj==1
strparamset = 'PRIOR';
nSYM = nSYMprior;
xparam_prior = set_prior(estim_params_,M_,options_);
[xparam_prior, estim_params_]= set_prior(estim_params_,M_,options_);
M_ = set_all_parameters(xparam_prior,estim_params_,M_);
elseif jj==2
strparamset = 'CALIBRATION';

2
tests/analytic_derivatives/burnside_3_order_PertParamsDerivs.mod
View File

@ -113,7 +113,7 @@ stoch_simul(order=@{ORDER},k_order_solver,irf=0,drop=0,periods=0,nograph);
identification(order=@{ORDER},nograph,no_identification_strength);
%make sure everything is computed at prior mean
xparam_prior = set_prior(estim_params_,M_,options_);
[xparam_prior, estim_params_]= set_prior(estim_params_,M_,options_);
M_ = set_all_parameters(xparam_prior,estim_params_,M_);
[oo_.dr,info,M_.params] = resol(0,M_, options_, oo_.dr, oo_.steady_state, oo_.exo_steady_state, oo_.exo_det_steady_state);

8
tests/gsa/cod_ML_morris/cod_ML_morris.mod
View File

@ -1,6 +1,6 @@
var c, y_ro, psi, s, tt, pi_h, mc_ro, i_ro, pi_ro, pi_eu, pi_f, y_eu, mc_eu, i_eu, g, a, cp, bnr, gs, as ;
varexo eps_a, eps_g, eps_cp, eps_s, eps_m, eps_as, eps_gs, eps_ms;
parameters alfa, niu, delt, teta_h, teta_f, bet, fi, sigm, h, ro, psi_pi, tetas, rhos, psi_pis, psi_y, psi_deltay, rho_g, rho_a, rho_cp, rho_s, rho_gs, rho_as ;
var c, y_ro, psi, s, tt, pi_h ${\pi_h}$, mc_ro, i_ro, pi_ro ${\pi_\rho}$, pi_eu, pi_f, y_eu, mc_eu, i_eu, g, a, cp, bnr, gs, as ;
varexo eps_a, eps_g, eps_cp, eps_s, eps_m $\varepsilon_m$, eps_as, eps_gs, eps_ms;
parameters alfa, niu, delt, teta_h, teta_f, bet, fi, sigm, h, ro, psi_pi $\psi_\pi$, tetas, rhos, psi_pis, psi_y, psi_deltay, rho_g, rho_a, rho_cp, rho_s, rho_gs, rho_as ;
alfa = 0.425 ;
bet = 0.99 ;
@ -121,7 +121,7 @@ stderr eps_as, 1, 0.01, 10 ;
end;
varobs y_ro, pi_ro, i_ro, s, y_eu, pi_eu, i_eu, tt ;
options_.TeX =1;
dynare_sensitivity (identification=1, nsam = 500, lik_only = 1, morris=2) ;
stoch_simul(order=2,irf=20) y_ro pi_ro i_ro s ;

73
tests/gsa/ls2003.mod
View File

@ -1,45 +1,4 @@
var y y_s R pie dq pie_s de A y_obs pie_obs R_obs ${R^{obs}}$;
varexo e_R ${e^R}$ e_q e_ys e_pies e_A;
parameters psi1 psi2 psi3 rho_R ${\rho_R}$ tau alpha rr k rho_q rho_A rho_ys rho_pies;
psi1 = 1.54;
psi2 = 0.25;
psi3 = 0.25;
rho_R = 0.5;
alpha = 0.3;
rr = 2.51;
k = 0.5;
tau = 0.5;
rho_q = 0.4;
rho_A = 0.2;
rho_ys = 0.9;
rho_pies = 0.7;
model(linear);
y = y(+1) - (tau +alpha*(2-alpha)*(1-tau))*(R-pie(+1))-alpha*(tau +alpha*(2-alpha)*(1-tau))*dq(+1) + alpha*(2-alpha)*((1-tau)/tau)*(y_s-y_s(+1))-A(+1);
pie = exp(-rr/400)*pie(+1)+alpha*exp(-rr/400)*dq(+1)-alpha*dq+(k/(tau+alpha*(2-alpha)*(1-tau)))*y+k*alpha*(2-alpha)*(1-tau)/(tau*(tau+alpha*(2-alpha)*(1-tau)))*y_s;
pie = de+(1-alpha)*dq+pie_s;
R = rho_R*R(-1)+(1-rho_R)*(psi1*pie+psi2*(y+alpha*(2-alpha)*((1-tau)/tau)*y_s)+psi3*de)+e_R;
dq = rho_q*dq(-1)+e_q;
y_s = rho_ys*y_s(-1)+e_ys;
pie_s = rho_pies*pie_s(-1)+e_pies;
A = rho_A*A(-1)+e_A;
y_obs = y-y(-1)+A;
pie_obs = 4*pie;
R_obs = 4*R;
end;
shocks;
var e_R = 1.25^2;
var e_q = 2.5^2;
var e_A = 1.89;
var e_ys = 1.89;
var e_pies = 1.89;
end;
varobs y_obs R_obs pie_obs dq de;
@#include "ls2003_model.inc"
estimated_params;
psi1 , gamma_pdf,1.5,0.5;
@ -66,7 +25,7 @@ options_.TeX=true;
disp(' ');
disp('NOW I DO STABILITY MAPPING and prepare sample for Reduced form Mapping');
disp(' ');
disp('Press ENTER to continue'); pause(5);
% disp('Press ENTER to continue'); pause(5);
dynare_sensitivity(redform=1,nodisplay,Nsam=512); //create sample of reduced form coefficients
// NOTE: since namendo is empty by default,
@ -75,7 +34,7 @@ dynare_sensitivity(redform=1,nodisplay,Nsam=512); //create sample of reduced for
disp(' ');
disp('ANALYSIS OF REDUCED FORM COEFFICIENTS');
disp(' ');
disp('Press ENTER to continue'); pause(5);
% disp('Press ENTER to continue'); pause(5);
dynare_sensitivity(nodisplay, load_stab=1, // load previously generated sample analysed for stability
redform=1, // do the reduced form mapping
@ -87,6 +46,7 @@ stab=0, // don't repeat again the stability mapping
Nsam=512);
disp(' ');
disp('THE PREVIOUS TWO CALLS COULD BE DONE TOGETHER');
disp('BY USING THE COMBINED CALL');
@ -94,7 +54,7 @@ disp(' ');
disp('dynare_sensitivity(redform=1,')
disp('threshold_redform=[-1 0], namendo=(pie,R), namexo=(e_R), namlagendo=(R));')
disp(' ');
disp('Press ENTER to continue'); pause(5);
% disp('Press ENTER to continue'); pause(5);
//dynare_sensitivity(
//redform=1, //create sample of reduced form coefficients
//logtrans_redform=1, // estimate log-transformed reduced form coefficients (default=0)
@ -107,7 +67,7 @@ disp('Press ENTER to continue'); pause(5);
disp(' ');
disp('MC FILTERING(rmse=1), TO MAP THE FIT FROM PRIORS');
disp('Press ENTER to continue'); pause(5);
%disp('Press ENTER to continue'); pause(5);
dynare_sensitivity(nodisplay, datafile='data_ca1.m',first_obs=8,nobs=79,prefilter=1, // also presample=2,loglinear, are admissible
load_stab=1, // load prior sample
@ -125,7 +85,7 @@ disp('logtrans_redform=1, namendo=(pie,R), namexo=(e_R), namlagendo=(R),')
disp('datafile=data_ca1.m,first_obs=8,nobs=79,prefilter=1,')
disp('istart_rmse=2, rmse=1);')
disp(' ');
disp('Press ENTER to continue'); pause(5);
% disp('Press ENTER to continue'); pause(5);
//dynare_sensitivity(
//redform=1, //create sample of reduced form coefficients
//logtrans_redform=1, // estimate log-transformed reduced form coefficients (default=0)
@ -142,14 +102,13 @@ disp('Press ENTER to continue'); pause(5);
disp(' ');
disp('I ESTIMATE THE MODEL');
disp(' ');
disp('Press ENTER to continue'); pause(5);
% disp('Press ENTER to continue'); pause(5);
// run this to generate posterior mode and Metropolis files if not yet done
estimation(datafile='data_ca1.m',first_obs=8,nobs=79,mh_nblocks=1,
prefilter=1,mh_jscale=0.5,mh_replic=5000,silent_optimizer, mode_compute=4, mh_drop=0.6, nodisplay,
bayesian_irf, filtered_vars, smoother) y_obs R_obs pie_obs dq de;
// run this to produce posterior samples of filtered, smoothed and irf variables, if not yet done
//estimation(datafile='data_ca1.m',first_obs=8,nobs=79,mh_nblocks=2,prefilter=1,mh_jscale=0.3,
// mh_replic=0, mode_file='ls2003/Output/ls2003_mode', mode_compute=0, load_mh_file, bayesian_irf,
@ -160,7 +119,7 @@ disp('WE DO STABILITY MAPPING AGAIN, EXPLORING AROUND THE MODE BY A WIDTH +- 0.2
disp('THEN WE DO RMSE analysis in this neighborhood');
disp('Typical for ML estimation, also feasible for posterior mode');
disp(' ');
disp('Press ENTER to continue'); pause(5);
% disp('Press ENTER to continue'); pause(5);
dynare_sensitivity(nodisplay, pprior=0,Nsam=512,neighborhood_width=0.2,
mode_file='ls2003/Output/ls2003_mode', // specifies the mode file where the mode and Hessian are stored
@ -171,7 +130,7 @@ disp(' ');
disp('WE DO STABILITY MAPPING AGAIN, BUT FOR MULTIVARIATE SAMPLE AT THE POSTERIOR MODE (or ML) and Hessian (pprior=0 & ppost=0)');
disp('Typical for ML estimation, also feasible for posterior mode');
disp(' ');
disp('Press ENTER to continue'); pause(5);
% disp('Press ENTER to continue'); pause(5);
dynare_sensitivity(nodisplay, pprior=0,Nsam=512,
mode_file='ls2003/Output/ls2003_mode' // specifies the mode file where the mode and Hessian are stored
@ -181,7 +140,7 @@ mode_file='ls2003/Output/ls2003_mode' // specifies the mode file where the mode
disp(' ');
disp('RMSE ANALYSIS FOR MULTIVARIATE SAMPLE AT THE POSTERIOR MODE');
disp(' ');
disp('Press ENTER to continue'); pause(5);
disp('Press ENTER to continue'); % pause(5);
dynare_sensitivity(nodisplay, mode_file='ls2003/Output/ls2003_mode',
datafile='data_ca1.m',first_obs=8,nobs=79,prefilter=1,
pprior=0,
@ -199,10 +158,10 @@ disp('dynare_sensitivity(pprior=0,Nsam=512,alpha2_stab=0.4,mode_file=ls2003_mode
disp('datafile=data_ca1.m,first_obs=8,nobs=79,prefilter=1,')
disp('rmse=1, alpha2_rmse=0, alpha_rmse=0);')
disp(' ');
disp('Press ENTER to continue'); pause(5);
% disp('Press ENTER to continue'); pause(5);
//dynare_sensitivity(pprior=0,Nsam=2048,alpha2_stab=0.4,mode_file='ls2003/Output/ls2003_mode',
//datafile='data_ca1.m',first_obs=8,nobs=79,prefilter=1,
//rmse=1,
//rmse=1pae
//alpha2_rmse=0, // no correlation analysis
//alpha_rmse=0 // no Smirnov sensitivity analysis
//);
@ -210,7 +169,7 @@ disp('Press ENTER to continue'); pause(5);
disp(' ');
disp('RMSE ANALYSIS FOR POSTERIOR MCMC sample (ppost=1)');
disp('Needs a call to dynare_estimation to load all MH environment');
disp('Press ENTER to continue'); pause(5);
% disp('Press ENTER to continue'); pause(5);
//estimation(datafile='data_ca1.m',first_obs=8,nobs=79,mh_nblocks=2, mode_file='ls2003/Output/ls2003_mode', load_mh_file,
// prefilter=1,mh_jscale=0.5,mh_replic=0, mode_compute=0, mh_drop=0.6);
@ -220,6 +179,8 @@ alpha2_rmse=0, // no correlation analysis
rmse=1,ppost=1,Nsam=512);
collect_latex_files;
if system(['pdflatex -halt-on-error -interaction=batchmode ' M_.fname '_TeX_binder.tex'])
[status, cmdout]=system(['pdflatex -halt-on-error -interaction=nonstopmode ' M_.fname '_TeX_binder.tex']);
if status
cmdout
error('TeX-File did not compile.')
end

70
tests/gsa/ls2003_model.inc Normal file
View File

@ -0,0 +1,70 @@
var y
y_s
R
pie $\pi$
dq
pie_s
de
A
y_obs ${y^{obs}}$
pie_obs ${\pi^{obs}}$
R_obs ${R^{obs}}$;
varexo e_R ${\varepsilon^R}$
e_q ${\varepsilon^q}$
e_ys ${\varepsilon^{ys}}$
e_pies ${\varepsilon^\pi}$
e_A ${\varepsilon^A}$;
parameters psi1 ${\psi_1}$
psi2 ${\psi_2}$
psi3 ${\psi_3}$
rho_R ${\rho_R}$
tau ${\tau}$
alpha ${\alpha}$
rr
k
rho_q ${\rho_q}$
rho_A ${\rho_A}$
rho_ys ${\rho_{ys}}$
rho_pies ${\rho_\pi}$;
psi1 = 1.54;
psi2 = 0.25;
psi3 = 0.25;
rho_R = 0.5;
alpha = 0.3;
rr = 2.51;
k = 0.5;
tau = 0.5;
rho_q = 0.4;
rho_A = 0.2;
rho_ys = 0.9;
rho_pies = 0.7;
model(linear);
y = y(+1) - (tau +alpha*(2-alpha)*(1-tau))*(R-pie(+1))-alpha*(tau +alpha*(2-alpha)*(1-tau))*dq(+1) + alpha*(2-alpha)*((1-tau)/tau)*(y_s-y_s(+1))-A(+1);
pie = exp(-rr/400)*pie(+1)+alpha*exp(-rr/400)*dq(+1)-alpha*dq+(k/(tau+alpha*(2-alpha)*(1-tau)))*y+k*alpha*(2-alpha)*(1-tau)/(tau*(tau+alpha*(2-alpha)*(1-tau)))*y_s;
pie = de+(1-alpha)*dq+pie_s;
R = rho_R*R(-1)+(1-rho_R)*(psi1*pie+psi2*(y+alpha*(2-alpha)*((1-tau)/tau)*y_s)+psi3*de)+e_R;
dq = rho_q*dq(-1)+e_q;
y_s = rho_ys*y_s(-1)+e_ys;
pie_s = rho_pies*pie_s(-1)+e_pies;
A = rho_A*A(-1)+e_A;
y_obs = y-y(-1)+A;
pie_obs = 4*pie;
R_obs = 4*R;
end;
shocks;
var e_R = 1.25^2;
var e_q = 2.5^2;
var e_A = 1.89;
var e_ys = 1.89;
var e_pies = 1.89;
end;
varobs y_obs R_obs pie_obs dq de;
options_.TeX=1;

43
tests/gsa/ls2003a.mod
View File

@ -1,45 +1,4 @@
var y y_s R pie dq pie_s de A y_obs pie_obs R_obs;
varexo e_R e_q e_ys e_pies e_A;
parameters psi1 psi2 psi3 rho_R tau alpha rr k rho_q rho_A rho_ys rho_pies;
psi1 = 1.54;
psi2 = 0.25;
psi3 = 0.25;
rho_R = 0.5;
alpha = 0.3;
rr = 2.51;
k = 0.5;
tau = 0.5;
rho_q = 0.4;
rho_A = 0.2;
rho_ys = 0.9;
rho_pies = 0.7;
model(linear);
y = y(+1) - (tau +alpha*(2-alpha)*(1-tau))*(R-pie(+1))-alpha*(tau +alpha*(2-alpha)*(1-tau))*dq(+1) + alpha*(2-alpha)*((1-tau)/tau)*(y_s-y_s(+1))-A(+1);
pie = exp(-rr/400)*pie(+1)+alpha*exp(-rr/400)*dq(+1)-alpha*dq+(k/(tau+alpha*(2-alpha)*(1-tau)))*y+k*alpha*(2-alpha)*(1-tau)/(tau*(tau+alpha*(2-alpha)*(1-tau)))*y_s;
pie = de+(1-alpha)*dq+pie_s;
R = rho_R*R(-1)+(1-rho_R)*(psi1*pie+psi2*(y+alpha*(2-alpha)*((1-tau)/tau)*y_s)+psi3*de)+e_R;
dq = rho_q*dq(-1)+e_q;
y_s = rho_ys*y_s(-1)+e_ys;
pie_s = rho_pies*pie_s(-1)+e_pies;
A = rho_A*A(-1)+e_A;
y_obs = y-y(-1)+A;
pie_obs = 4*pie;
R_obs = 4*R;
end;
shocks;
var e_R = 1.25^2;
var e_q = 2.5^2;
var e_A = 1.89;
var e_ys = 1.89;
var e_pies = 1.89;
end;
varobs y_obs R_obs pie_obs dq de;
@#include "ls2003_model.inc"
estimated_params;
psi1 , gamma_pdf,1.5,0.5;

56
tests/gsa/ls2003ide.mod
View File

@ -1,45 +1,4 @@
var y y_s R pie dq pie_s de A y_obs pie_obs R_obs;
varexo e_R e_q e_ys e_pies e_A;
parameters psi1 psi2 psi3 rho_R tau alpha rr k rho_q rho_A rho_ys rho_pies;
psi1 = 1.54;
psi2 = 0.25;
psi3 = 0.25;
rho_R = 0.5;
alpha = 0.3;
rr = 2.51;
k = 0.5;
tau = 0.5;
rho_q = 0.4;
rho_A = 0.2;
rho_ys = 0.9;
rho_pies = 0.7;
model(linear);
y = y(+1) - (tau +alpha*(2-alpha)*(1-tau))*(R-pie(+1))-alpha*(tau +alpha*(2-alpha)*(1-tau))*dq(+1) + alpha*(2-alpha)*((1-tau)/tau)*(y_s-y_s(+1))-A(+1);
pie = exp(-rr/400)*pie(+1)+alpha*exp(-rr/400)*dq(+1)-alpha*dq+(k/(tau+alpha*(2-alpha)*(1-tau)))*y+k*alpha*(2-alpha)*(1-tau)/(tau*(tau+alpha*(2-alpha)*(1-tau)))*y_s;
pie = de+(1-alpha)*dq+pie_s;
R = rho_R*R(-1)+(1-rho_R)*(psi1*pie+psi2*(y+alpha*(2-alpha)*((1-tau)/tau)*y_s)+psi3*de)+e_R;
dq = rho_q*dq(-1)+e_q;
y_s = rho_ys*y_s(-1)+e_ys;
pie_s = rho_pies*pie_s(-1)+e_pies;
A = rho_A*A(-1)+e_A;
y_obs = y-y(-1)+A;
pie_obs = 4*pie;
R_obs = 4*R;
end;
shocks;
var e_R = 1.25^2;
var e_q = 2.5^2;
var e_A = 1.89;
var e_ys = 1.89;
var e_pies = 1.89;
end;
varobs y_obs R_obs pie_obs dq de;
@#include "ls2003_model.inc"
estimated_params;
psi1 , gamma_pdf,1.5,0.5;
@ -65,15 +24,14 @@ end;
disp('CREATE SCREENING SAMPLE, CHECK FOR STABILITY AND PERFORM A SCREENING FOR IDENTIFICATION ANALYSIS');
disp('TYPE II ERRORS')
disp(' ')
disp('PRESS ENTER TO CONTUNUE');
pause(5);
% disp('PRESS ENTER TO CONTUNUE');
% pause(5);
dynare_sensitivity(identification=1, morris_nliv=6, morris_ntra=50);
disp('CREATE MC SAMPLE, CHECK FOR STABILITY AND PERFORM IDENTIFICATION ANALYSIS');
disp('WIth analytic derivatives')
disp('With analytic derivatives')
disp(' ')
disp('PRESS ENTER TO CONTUNUE');
pause(5);
dynare_sensitivity(identification=1, morris=2);
% disp('PRESS ENTER TO CONTUNUE');
% pause(5);
dynare_sensitivity(identification=1, morris=2);

47
tests/gsa/ls2003scr.mod
View File

@ -1,45 +1,4 @@
var y y_s R pie dq pie_s de A y_obs pie_obs R_obs;
varexo e_R e_q e_ys e_pies e_A;
parameters psi1 psi2 psi3 rho_R tau alpha rr k rho_q rho_A rho_ys rho_pies;
psi1 = 1.54;
psi2 = 0.25;
psi3 = 0.25;
rho_R = 0.5;
alpha = 0.3;
rr = 2.51;
k = 0.5;
tau = 0.5;
rho_q = 0.4;
rho_A = 0.2;
rho_ys = 0.9;
rho_pies = 0.7;
model(linear);
y = y(+1) - (tau +alpha*(2-alpha)*(1-tau))*(R-pie(+1))-alpha*(tau +alpha*(2-alpha)*(1-tau))*dq(+1) + alpha*(2-alpha)*((1-tau)/tau)*(y_s-y_s(+1))-A(+1);
pie = exp(-rr/400)*pie(+1)+alpha*exp(-rr/400)*dq(+1)-alpha*dq+(k/(tau+alpha*(2-alpha)*(1-tau)))*y+k*alpha*(2-alpha)*(1-tau)/(tau*(tau+alpha*(2-alpha)*(1-tau)))*y_s;
pie = de+(1-alpha)*dq+pie_s;
R = rho_R*R(-1)+(1-rho_R)*(psi1*pie+psi2*(y+alpha*(2-alpha)*((1-tau)/tau)*y_s)+psi3*de)+e_R;
dq = rho_q*dq(-1)+e_q;
y_s = rho_ys*y_s(-1)+e_ys;
pie_s = rho_pies*pie_s(-1)+e_pies;
A = rho_A*A(-1)+e_A;
y_obs = y-y(-1)+A;
pie_obs = 4*pie;
R_obs = 4*R;
end;
shocks;
var e_R = 1.25^2;
var e_q = 2.5^2;
var e_A = 1.89;
var e_ys = 1.89;
var e_pies = 1.89;
end;
varobs y_obs R_obs pie_obs dq de;
@#include "ls2003_model.inc"
estimated_params;
psi1 , gamma_pdf,1.5,0.5;
@ -63,8 +22,8 @@ end;
disp('CREATE SCREENING SAMPLE, CHECK FOR STABILITY AND PERFORM SENSITIVITY ANALYSIS');
disp('PRESS ENTER TO CONTUNUE');
pause;
% disp('PRESS ENTER TO CONTUNUE');
% pause;
dynare_sensitivity(morris=1, morris_nliv=6, morris_ntra=20, redform=1,
namendo=(:), namexo=(:), namlagendo=(:));

10
tests/identification/as2007/as2007.mod
View File

@ -19,9 +19,9 @@
% along with Dynare. If not, see <https://www.gnu.org/licenses/>.
% =========================================================================
var pie y R g z YGR INFL INT;
varexo e_R e_g e_z;
parameters tau kap psi1 psi2 rhoR rhog rhoz rr_steady pi_steady gam_steady std_R std_g std_z;
var pie $\pi$ y R g z YGR INFL INT;
varexo e_R $\varepsilon_R$ e_g e_z;
parameters tau $\tau$ kap $\kappa$ psi1 psi2 rhoR rhog rhoz rr_steady pi_steady gam_steady std_R std_g std_z;
tau = 2;
kap = 0.15;
@ -96,7 +96,9 @@ identification(tex);
identification(tex,advanced=1,max_dim_cova_group=3,prior_mc=250,no_identification_spectrum, no_identification_minimal);
collect_latex_files;
if system(['pdflatex -halt-on-error -interaction=batchmode ' M_.fname '_TeX_binder.tex'])
[status, cmdout]=system(['pdflatex -halt-on-error -interaction=nonstopmode ' M_.fname '_TeX_binder.tex']);
if status
cmdout
error('TeX-File did not compile.')
end
close all;

4
tests/identification/correlated_errors/fs2000_corr.mod
View File

@ -134,6 +134,8 @@ identification(advanced=1,max_dim_cova_group=3,prior_mc=250,tex);
collect_latex_files;
if system(['pdflatex -halt-on-error -interaction=batchmode ' M_.fname '_TeX_binder.tex'])
[status, cmdout]=system(['pdflatex -halt-on-error -interaction=nonstopmode ' M_.fname '_TeX_binder.tex']);
if status
cmdout
error('TeX-File did not compile.')
end

2
tests/identification/forward_looking/forward_looking_empty_ghx.mod
View File

@ -26,7 +26,7 @@ TRUE_SOLUTION2 = 1/(KAPPA*PSI/TAU +1)*[1 KAPPA*PSI PSI;
-1/TAU 1 -PSI/TAU;
-KAPPA/TAU KAPPA 1];
% note that BETA drops out from the solution
stoch_simul(order=1,noprint,irf=0,nomoments);
if max(max(abs(TRUE_SOLUTION1 - oo_.dr.ghu))) > 1e-15
error('Something wrong with perturbation');
end

1
tests/identification/kim/kim2.mod
View File

@ -125,4 +125,5 @@ end
% Integration test if identification works without priors
estim_params_=[];
dumpy=0;
identification(advanced=1,max_dim_cova_group=3);

6
tests/optimal_policy/Ramsey/Ramsey_Example_estimation.mod
View File

@ -245,6 +245,8 @@ write_latex_original_model(write_equation_tags);
write_latex_steady_state_model;
collect_latex_files;
if system(['pdflatex -halt-on-error -interaction=batchmode ' M_.fname '_TeX_binder.tex'])
[status, cmdout]=system(['pdflatex -halt-on-error -interaction=nonstopmode ' M_.fname '_TeX_binder.tex']);
if status
cmdout
error('TeX-File did not compile.')
end
end

4
tests/shock_decomposition/ls2003_plot.mod
View File

@ -165,6 +165,8 @@ realtime_shock_decomposition(fast_realtime=75) y_obs R_obs pie_obs dq de;
squeeze_shock_decomposition;
collect_latex_files;
if system(['pdflatex -halt-on-error -interaction=batchmode ' M_.fname '_TeX_binder.tex'])
[status, cmdout]=system(['pdflatex -halt-on-error -interaction=nonstopmode ' M_.fname '_TeX_binder.tex']);
if status
cmdout
error('TeX-File did not compile.')
end