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
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
|
function oo_recursive_=dynare_estimation(var_list,dname)
% function dynare_estimation(var_list, dname)
% runs the estimation of the model
%
% INPUTS
% var_list: selected endogenous variables vector
% dname: alternative directory name
%
% OUTPUTS
% oo_recursive_: cell array containing the results structures from recursive estimation
%
% SPECIAL REQUIREMENTS
% none
% Copyright (C) 2003-2020 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/>.
global options_ oo_ M_ dataset_ dataset_info
oo_recursive_={};
mode_file0 = options_.mode_file; % store mode_file set by the user
% Test if the order of approximation is nonzero (the preprocessor tests if order is non negative).
if isequal(options_.order,0)
error('Estimation:: The order of the Taylor approximation cannot be 0!')
end
% Decide if a DSGE or DSGE-VAR has to be estimated.
if ~isempty(strmatch('dsge_prior_weight',M_.param_names))
options_.dsge_var = 1;
end
var_list = check_list_of_variables(options_, M_, var_list);
options_.varlist = var_list;
nobs = sort(options_.nobs);
first_obs = sort(options_.first_obs);
nnobs = length(nobs);
nfirstobs = length(first_obs);
if nnobs~=1 && nfirstobs~=1
error('You cannot simultaneously do rolling window and recursive estimation')
end
horizon = options_.forecast;
if nargin<2 || ~exist('dname','var') || isempty(dname)
dname = options_.dirname;
end
M_.dname = dname;
if (isnumeric(options_.mode_compute) && options_.mode_compute && options_.analytic_derivation) ... %no user supplied function
|| (~isnumeric(options_.mode_compute) && options_.analytic_derivation) % user supplied function
analytic_derivation0=options_.analytic_derivation;
options_.analytic_derivation=1;
end
if options_.logged_steady_state
oo_.dr.ys=exp(oo_.dr.ys);
oo_.steady_state=exp(oo_.steady_state);
options_.logged_steady_state=0;
end
if nnobs>1 || nfirstobs > 1
for i=1:max(nnobs,nfirstobs)
if nnobs>1
options_.nobs = nobs(i);
M_.dname = [dname '_' int2str(nobs(i))];
elseif nfirstobs>1
options_.first_obs=first_obs(i);
M_.dname = [dname '_' int2str(first_obs(i))];
end
dynare_estimation_1(var_list,M_.dname);
if isequal(i,1) && options_.mode_compute ~= 0
options_.mode_file = [M_.fname '_mode'];
end
if options_.recursive_estimation_restart
for j=1:options_.recursive_estimation_restart
dynare_estimation_1(var_list,M_.dname);
end
end
if nnobs>1
oo_recursive_{nobs(i)} = oo_;
elseif nfirstobs>1
oo_recursive_{first_obs(i)} = oo_;
end
end
else
dynare_estimation_1(var_list,dname);
end
if isnumeric(options_.mode_compute) && options_.mode_compute && options_.analytic_derivation
options_.analytic_derivation=analytic_derivation0;
end
if nnobs > 1 && horizon > 0
mh_replic = options_.mh_replic;
endo_names = M_.endo_names;
n_varobs = length(options_.varobs);
if isempty(var_list)
var_list = endo_names;
nvar = length(endo_names);
SelecVariables = transpose(1:nvar);
else
nvar = length(var_list);
SelecVariables = [];
for i=1:nvar
if ~isempty(strmatch(var_list{i}, endo_names, 'exact'))
SelecVariables = [SelecVariables; strmatch(var_list{i}, endo_names, 'exact')];
else
error(['Estimation:: ' var_list{i} ' isn''t an endogenous variable'])
end
end
end
IdObs = zeros(n_varobs,1);
for j=1:n_varobs
iobs = strmatch(options_.varobs{j},var_list,'exact');
if ~isempty(iobs)
IdObs(j,1) = iobs;
end
end
gend = dataset_.nobs;
time_offset=min(3,gend-1); %for observables, plot 3 previous periods unless data is shorter
k = time_offset+min(nobs(end)-nobs(1)+horizon, ...
size(dataset_.data,1)-nobs(1));
data2 = dataset_info.rawdata(end-k+1:end,:);
[nbplt,nr,nc,lr,lc,nstar] = pltorg(nvar);
m = 1;
plot_index=0;
OutputDirectoryName = CheckPath('graphs',M_.fname);
for i = 1:length(var_list)
if mod(i,nstar) == 1
plot_index=plot_index+1;
hfig = dyn_figure(options_.nodisplay,'Name',['Out of sample forecasts (',num2str(plot_index),')']);
m = 1;
end
subplot(nr,nc,m)
hold on
if any(i==IdObs)
k2 = find(i==IdObs);
offsetx = 3;
plot(nobs(1)-offsetx+1:nobs(1)-offsetx+k,data2(end-k+1:end,k2)','-k','linewidth',2);
else
offsetx = 0;
end
vname = var_list{i};
for j=1:nnobs
if mh_replic > 0
oo_.RecursiveForecast.Mean.(vname)(j,:) = ...
oo_recursive_{nobs(j)}.MeanForecast.Mean.(vname);
oo_.RecursiveForecast.HPDinf.(vname)(j,:) = ...
oo_recursive_{nobs(j)}.MeanForecast.HPDinf.(vname);
oo_.RecursiveForecast.HPDsup.(vname)(j,:) = ...
oo_recursive_{nobs(j)}.MeanForecast.HPDsup.(vname);
oo_.RecursiveForecast.HPDTotalinf.(vname)(j,:) = ...
oo_recursive_{nobs(j)}.PointForecast.HPDinf.(vname);
oo_.RecursiveForecast.HPDTotalsup.(vname)(j,:) = ...
oo_recursive_{nobs(j)}.PointForecast.HPDsup.(vname);
else
oo_.RecursiveForecast.Mean.(vname)(j,:) =...
oo_recursive_{nobs(j)}.forecast.Mean.(vname);
oo_.RecursiveForecast.HPDinf.(vname)(j,:) =...
oo_recursive_{nobs(j)}.forecast.HPDinf.(vname);
oo_.RecursiveForecast.HPDsup.(vname)(j,:) =...
oo_recursive_{nobs(j)}.forecast.HPDsup.(vname);
end
x = nobs(1)+nobs(j)-nobs(1)+(1:horizon);
y = oo_.RecursiveForecast.Mean.(vname)(j,:);
y1 = oo_.RecursiveForecast.HPDinf.(vname)(j,:);
y2 = oo_.RecursiveForecast.HPDsup.(vname)(j,:);
plot(x,y,'-b','linewidth',2)
plot(x,y1,'--g', ...
'linewidth',1.5)
plot(x,y2,'--g', ...
'linewidth',1.5)
if mh_replic
y3 = oo_.RecursiveForecast.HPDTotalinf.(vname)(j,:);
y4 = oo_.RecursiveForecast.HPDTotalsup.(vname)(j,:);
plot(x,y3,'--r', ...
'linewidth',1.5)
plot(x,y4,'--r','linewidth',1.5)
end
end
box on
title(vname,'Interpreter','none')
hold off
xlim([nobs(1)-offsetx nobs(end)+horizon])
m = m + 1;
if mod(i+1,nstar) == 1 || i==length(var_list)
dyn_saveas(hfig,[M_.fname,filesep,'graphs',filesep M_.fname '_RecursiveForecasts_' int2str(plot_index)],options_.nodisplay,options_.graph_format);
end
end
end
options_.mode_file = mode_file0;
%reset stored mode-file to user defined one (and in case it was only set by the recursive estimation)
oo_.gui.ran_estimation = true;
|
