Skip to content
Snippets Groups Projects
Select Git revision
  • 6072aa0b462484da990ffa00ee39fb141fa9513f
  • master default
  • lto
  • pf_block_problem
  • r2025a
  • asm
  • kronecker
  • 4.5.6
  • 4.5.5
  • 4.5.4
  • 4.5.3
  • 4.5.2
  • 4.5.1
  • 4.5.0
  • 4.4.3
  • 4.4.2
  • 4.4.1
  • 4.4.0
  • 4.4-beta1
  • 4.3.3
  • 4.3.2
  • 4.3.1
  • 4.3.0
  • 4.2.5
  • 4.2.4
  • 4.2.3
  • 4.2.2
27 results

correlation_mc_analysis.m

Blame
  • Forked from Dynare / dynare
    8125 commits behind the upstream repository.
    Johannes Pfeifer's avatar
    Johannes Pfeifer authored
    Necessary after switching to indirect indexing of structure fields
    1f18a246
    History
    correlation_mc_analysis.m 6.27 KiB
    function oo_ = correlation_mc_analysis(SampleSize,type,dname,fname,vartan,nvar,var1,var2,nar,mh_conf_sig,oo_,M_,options_)
    % This function analyses the (posterior or prior) distribution of the
    % endogenous variables correlation function.
    
    % Copyright (C) 2008-2013 Dynare Team
    %
    % This file is part of Dynare.
    %
    % Dynare is free software: you can redistribute it and/or modify
    % it under the terms of the GNU General Public License as published by
    % the Free Software Foundation, either version 3 of the License, or
    % (at your option) any later version.
    %
    % Dynare is distributed in the hope that it will be useful,
    % but WITHOUT ANY WARRANTY; without even the implied warranty of
    % MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    % GNU General Public License for more details.
    %
    % You should have received a copy of the GNU General Public License
    % along with Dynare.  If not, see <http://www.gnu.org/licenses/>.
    
    if strcmpi(type,'posterior')
        TYPE = 'Posterior';
        PATH = [dname '/metropolis/'];
    else
        TYPE = 'Prior';
        PATH = [dname '/prior/moments/'];
    end
    
    indx1 = check_name(vartan,var1);
    if isempty(indx1)
        disp([ type '_analysis:: ' var1 ' is not a stationary endogenous variable!'])
        return
    end
    if ~isempty(var2)
        indx2 = check_name(vartan,var2);
        if isempty(indx2)
            disp([ type '_analysis:: ' var2 ' is not a stationary endogenous variable!'])
            return
        end
    else
        indx2 = indx1;
        var2 = var1;
    end
    
    var1=deblank(var1);
    var2=deblank(var2);
    
    if isfield(oo_,[TYPE 'TheoreticalMoments'])
        temporary_structure = oo_.([TYPE, 'TheoreticalMoments']);
        if isfield(temporary_structure,'dsge')
            temporary_structure = oo_.([TYPE, 'TheoreticalMoments']).dsge;
            if isfield(temporary_structure,'correlation')
                temporary_structure = oo_.([TYPE, 'TheoreticalMoments']).dsge.correlation.Mean;
                if isfield(temporary_structure,deblank(var1))
                    temporary_structure_1 = oo_.([TYPE, 'TheoreticalMoments']).dsge.correlation.Mean.(var1);
                    if isfield(temporary_structure_1,deblank(var2))
                        temporary_structure_2 = temporary_structure_1.(var2);
                        l1 = length(temporary_structure_2);
                        if l1<nar
                            % INITIALIZATION:
                            oo_ = initialize_output_structure(var1,var2,nar,type,oo_);
                            delete([PATH fname '_' TYPE 'Correlations*'])
                            [nvar,vartan,NumberOfFiles] = ...
                                dsge_simulated_theoretical_correlation(SampleSize,nar,M_,options_,oo_,type);
                        else
                            if ~isnan(temporary_structure_2(nar))
                                %Nothing to do.
                                return
                            end
                        end
                    else
                        oo_ = initialize_output_structure(var1,var2,nar,TYPE,oo_,options_);
                    end
                else
                    oo_ = initialize_output_structure(var1,var2,nar,TYPE,oo_,options_);
                end
            else
                oo_ = initialize_output_structure(var1,var2,nar,TYPE,oo_,options_);
            end
        else
            oo_ = initialize_output_structure(var1,var2,nar,TYPE,oo_,options_);
        end
    else
        oo_ = initialize_output_structure(var1,var2,nar,TYPE,oo_,options_);
    end
    ListOfFiles = dir([ PATH  fname '_' TYPE 'Correlations*.mat']);
    i1 = 1; tmp = zeros(SampleSize,1);
    for file = 1:length(ListOfFiles)
        load([ PATH  ListOfFiles(file).name ]);
        i2 = i1 + rows(Correlation_array) - 1;
        tmp(i1:i2) = Correlation_array(:,indx1,indx2,nar);
        i1 = i2+1;
    end
    name = [ var1 '.' var2 ];
    if options_.estimation.moments_posterior_density.indicator
        [p_mean, p_median, p_var, hpd_interval, p_deciles, density] = ...
            posterior_moments(tmp,1,mh_conf_sig);
    else
        [p_mean, p_median, p_var, hpd_interval, p_deciles] = ...
            posterior_moments(tmp,0,mh_conf_sig);
    end
    if isfield(oo_,[ TYPE 'TheoreticalMoments'])
        temporary_structure = oo_.([TYPE, 'TheoreticalMoments']);
        if isfield(temporary_structure,'dsge')
            temporary_structure = oo_.([TYPE, 'TheoreticalMoments']).dsge;
            if isfield(temporary_structure,'correlation')
                oo_ = fill_output_structure(var1,var2,TYPE,oo_,'Mean',nar,p_mean);
                oo_ = fill_output_structure(var1,var2,TYPE,oo_,'Median',nar,p_median);
                oo_ = fill_output_structure(var1,var2,TYPE,oo_,'Variance',nar,p_var);
                oo_ = fill_output_structure(var1,var2,TYPE,oo_,'HPDinf',nar,hpd_interval(1));
                oo_ = fill_output_structure(var1,var2,TYPE,oo_,'HPDsup',nar,hpd_interval(2));
                oo_ = fill_output_structure(var1,var2,TYPE,oo_,'deciles',nar,p_deciles);
                if options_.estimation.moments_posterior_density.indicator
                    oo_ = fill_output_structure(var1,var2,TYPE,oo_,'density',nar,density);
                end
            end
        end
    end
    
    function oo_ = initialize_output_structure(var1,var2,nar,type,oo_,options_)
    oo_.([type, 'TheoreticalMoments']).dsge.correlation.Mean.(var1).(var2) = NaN(nar,1);
    oo_.([type, 'TheoreticalMoments']).dsge.correlation.Median.(var1).(var2) = NaN(nar,1);
    oo_.([type, 'TheoreticalMoments']).dsge.correlation.Variance.(var1).(var2) = NaN(nar,1);
    oo_.([type, 'TheoreticalMoments']).dsge.correlation.HPDinf.(var1).(var2) = NaN(nar,1);
    oo_.([type, 'TheoreticalMoments']).dsge.correlation.HPDsup.(var1).(var2) = NaN(nar,1);
    oo_.([type, 'TheoreticalMoments']).dsge.correlation.deciles.(var1).(var2) = cell(nar,1);
    if options_.estimation.moments_posterior_density.indicator
        oo_.([type, 'TheoreticalMoments']).dsge.correlation.density.(var1).(var2) = cell(nar,1);
    end
    for i=1:nar
        if options_.estimation.moments_posterior_density.indicator
            oo_.([type, 'TheoreticalMoments']).dsge.correlation.density.(var1).(var2)(i,1) = {NaN};
        end
        oo_.([type, 'TheoreticalMoments']).dsge.correlation.deciles.(var1).(var2)(i,1) = {NaN};
    end
    
    function oo_ = fill_output_structure(var1,var2,type,oo_,moment,lag,result)
    switch moment
      case {'Mean','Median','Variance','HPDinf','HPDsup'} 
        oo_.([type,  'TheoreticalMoments']).dsge.correlation.(moment).(var1).(var2)(lag,1) = result;
      case {'deciles','density'}
        oo_.([type, 'TheoreticalMoments']).dsge.correlation.(moment).(var1).(var2)(lag,1) = {result};
      otherwise
        disp('fill_output_structure:: Unknown field!')
    end