111 lines
3.3 KiB
Matlab
111 lines
3.3 KiB
Matlab
function r = get_posterior_timeseries(type, endo)
|
|
% Returns the posterior distribution of either
|
|
% smoothed/updated/filtered/forecast timeseries for a given endogenous variable.
|
|
% "type" must be either 'smoothed', 'updated', 'filtered' or 'forecast'.
|
|
%
|
|
% For filtered variables, returns a matrix with step-ahead in lines and periods
|
|
% in columns.
|
|
% For forecasts, returns the "point" forecast (i.e. with uncertainty about both
|
|
% parameters and shocks).
|
|
|
|
% Copyright © 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 <https://www.gnu.org/licenses/>.
|
|
|
|
global M_
|
|
|
|
endo_id = find(strcmp(endo, M_.endo_names));
|
|
|
|
if isempty(endo_id)
|
|
error('Unknown endogenous variable %s', endo)
|
|
end
|
|
|
|
r = struct('draw', [], type, []);
|
|
|
|
|
|
% Fetch parameter draws
|
|
|
|
DrawsFiles = dir([M_.dname '/metropolis/' M_.fname '_param*' ]);
|
|
NumberDrawsFiles = 0;
|
|
for i=1:length(DrawsFiles)
|
|
% We need to filter out the _param_irf* files
|
|
if ~isempty(regexp(DrawsFiles(i).name, '.*_param[0-9]+\.mat'))
|
|
NumberDrawsFiles = NumberDrawsFiles + 1;
|
|
end
|
|
end
|
|
if NumberDrawsFiles == 0
|
|
error('Can''t find posterior draws file(s)')
|
|
end
|
|
|
|
idx = 1;
|
|
for file = 1:NumberDrawsFiles
|
|
load([M_.dname '/metropolis/' M_.fname '_param' int2str(file) ],'stock');
|
|
for i = 1:size(stock, 1)
|
|
r(idx).draw = stock(i, :);
|
|
idx = idx+1;
|
|
end
|
|
end
|
|
|
|
|
|
% Fetch timeseries
|
|
|
|
switch type
|
|
case 'smoothed'
|
|
basename = 'smooth';
|
|
case 'updated'
|
|
basename = 'update';
|
|
case 'filtered'
|
|
basename = 'filter_step_ahead';
|
|
case 'forecast'
|
|
basename = 'forc_point';
|
|
otherwise
|
|
error('Unknown type requested. Should be one of: smoothed, updated, filtered, forecast')
|
|
end
|
|
|
|
|
|
if strcmp(type, 'smoothed')
|
|
SmoothedFiles = dir([M_.dname '/metropolis/' M_.fname '_smooth*']);
|
|
NumberTimeseriesFiles = 0;
|
|
for i=1:length(SmoothedFiles)
|
|
%% We need to filter out the _smoothed_{constant,trend}* files
|
|
if ~isempty(regexp(SmoothedFiles(i).name, '.*_smooth[0-9]+\.mat'))
|
|
NumberTimeseriesFiles = NumberTimeseriesFiles + 1;
|
|
end
|
|
end
|
|
else
|
|
NumberTimeseriesFiles = length(dir([M_.dname '/metropolis/' M_.fname '_' ...
|
|
basename '*']));
|
|
end
|
|
if NumberTimeseriesFiles == 0
|
|
error('Can''t find file(s) with posterior timeseries of requested type')
|
|
end
|
|
|
|
idx = 1;
|
|
for file = 1:NumberTimeseriesFiles
|
|
load([M_.dname '/metropolis/' M_.fname '_' basename int2str(file) ],'stock');
|
|
if strcmp(type, 'filtered')
|
|
for i = 1:size(stock, 4)
|
|
r(idx).(type) = reshape(stock(:, endo_id, :, i), size(stock,1), size(stock,3));
|
|
idx = idx+1;
|
|
end
|
|
else
|
|
for i = 1:size(stock, 3)
|
|
r(idx).(type) = stock(endo_id, :, i)';
|
|
idx = idx+1;
|
|
end
|
|
end
|
|
end
|