File: shock_decomposition.m

package info (click to toggle)
dynare 4.5.7-1
  • links: PTS, VCS
  • area: main
  • in suites: buster
  • size: 49,408 kB
  • sloc: cpp: 84,998; ansic: 29,058; pascal: 13,843; sh: 4,833; objc: 4,236; yacc: 3,622; makefile: 2,278; lex: 1,541; python: 236; lisp: 69; xml: 8
file content (127 lines) | stat: -rw-r--r-- 4,241 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
function [oo_,M_] = shock_decomposition(M_,oo_,options_,varlist,bayestopt_,estim_params_)
% function z = shock_decomposition(M_,oo_,options_,varlist)
% Computes shocks contribution to a simulated trajectory. The field set is
% oo_.shock_decomposition. It is a n_var by nshock+2 by nperiods array. The
% first nshock columns store the respective shock contributions, column n+1
% stores the role of the initial conditions, while column n+2 stores the
% value of the smoothed variables.  Both the variables and shocks are stored
% in the order of declaration, i.e. M_.endo_names and M_.exo_names, respectively.
%
% INPUTS
%    M_:          [structure]  Definition of the model
%    oo_:         [structure]  Storage of results
%    options_:    [structure]  Options
%    varlist:     [char]       List of variables
%    bayestopt_:  [structure]  describing the priors
%    estim_params_: [structure] characterizing parameters to be estimated
%
% OUTPUTS
%    oo_:         [structure]  Storage of results
%    M_:          [structure]  Definition of the model; makes sure that
%                               M_.params is correctly updated
%
% SPECIAL REQUIREMENTS
%    none

% Copyright (C) 2009-2017 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 <http://www.gnu.org/licenses/>.

% indices of endogenous variables
if size(varlist,1) == 0
    varlist = M_.endo_names(1:M_.orig_endo_nbr,:);
end

[i_var,nvar,index_uniques] = varlist_indices(varlist,M_.endo_names);
varlist=varlist(index_uniques,:);

% number of variables
endo_nbr = M_.endo_nbr;

% number of shocks
nshocks = M_.exo_nbr;

% parameter set
parameter_set = options_.parameter_set;
if isempty(parameter_set)
    if isfield(oo_,'posterior_mean')
        parameter_set = 'posterior_mean';
    elseif isfield(oo_,'mle_mode')
        parameter_set = 'mle_mode';
    elseif isfield(oo_,'posterior')
        parameter_set = 'posterior_mode';
    else
        error(['shock_decomposition: option parameter_set is not specified ' ...
               'and posterior mode is not available'])
    end
end


options_.selected_variables_only = 0; %make sure all variables are stored
options_.plot_priors=0;
[oo_, M_, junk1, junk2, Smoothed_Variables_deviation_from_mean] = evaluate_smoother(parameter_set, varlist, M_, oo_, options_, bayestopt_, estim_params_);

% reduced form
dr = oo_.dr;

% data reordering
order_var = dr.order_var;
inv_order_var = dr.inv_order_var;


% coefficients
A = dr.ghx;
B = dr.ghu;

% initialization
gend = size(oo_.SmoothedShocks.(deblank(M_.exo_names(1,:))),1);
epsilon=NaN(nshocks,gend);
for i=1:nshocks
    epsilon(i,:) = oo_.SmoothedShocks.(deblank(M_.exo_names(i,:)));
end

z = zeros(endo_nbr,nshocks+2,gend);

z(:,end,:) = Smoothed_Variables_deviation_from_mean;

maximum_lag = M_.maximum_lag;

k2 = dr.kstate(find(dr.kstate(:,2) <= maximum_lag+1),[1 2]);
i_state = order_var(k2(:,1))+(min(i,maximum_lag)+1-k2(:,2))*M_.endo_nbr;
for i=1:gend
    if i > 1 && i <= maximum_lag+1
        lags = min(i-1,maximum_lag):-1:1;
    end

    if i > 1
        tempx = permute(z(:,1:nshocks,lags),[1 3 2]);
        m = min(i-1,maximum_lag);
        tempx = [reshape(tempx,endo_nbr*m,nshocks); zeros(endo_nbr*(maximum_lag-i+1),nshocks)];
        z(:,1:nshocks,i) = A(inv_order_var,:)*tempx(i_state,:);
        lags = lags+1;
    end

    if i > options_.shock_decomp.init_state
        z(:,1:nshocks,i) = z(:,1:nshocks,i) + B(inv_order_var,:).*repmat(epsilon(:,i)',endo_nbr,1);
    end
    z(:,nshocks+1,i) = z(:,nshocks+2,i) - sum(z(:,1:nshocks,i),2);
end

oo_.shock_decomposition = z;

if ~options_.no_graph.shock_decomposition
    plot_shock_decomposition(M_,oo_,options_,varlist);
end