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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292
|
function get_ar_ec_matrices(model_name, model_type)
%function get_ar_ec_matrices(model_name, model_type)
%
% Returns the autoregressive matrix associated with the auxiliary model specified by
% model_name. Output is stored in cellarray oo_.(model_type).(model_name).ar,
% with oo_.(model_type).(model_name).ar(:,:,i) being the AR matrix at time t-i. Each
% AR matrix is stored with rows and columns organized by the ordering of the
% equation tags found in M_.(model_type).(model_name).eqtags.
% oo_.(model_type).(model_name).ec contains those entries that are not
% autoregressive.
%
% INPUTS
%
% model_name [string] the name of the auxiliary model
% model_type [string] the type of the auxiliary model ('var' or
% 'trend_component'. If not passed, the
% value is set by the function; if a 'var'
% subfield is found, that is used. Otherwise
% 'trend_component' is used (if it exists as
% a subfield of M_.
%
% OUTPUTS
%
% NONE
% 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 M_ oo_
%% Check inputs
assert(nargin <= 2, 'This function requires one or two arguments');
assert(~isempty(model_name) && ischar(model_name), ...
'The first argument must be a non-empty string');
if nargin < 2
model_type = 'var';
if ~(isfield(M_, model_type) && isfield(M_.(model_type), model_name))
model_type = 'trend_component';
if ~(isfield(M_, model_type) && isfield(M_.(model_type), model_name))
error(['Could not find ' model_name ' in M_.var or ' ...
'M_.trend_component. First declare it via the var_model ' ...
'or trend_component_model statement.']);
end
end
else
assert(~isempty(model_type) && ischar(model_type), ...
'If provided, the second argument must be a non-empty string');
if ~(isfield(M_, model_type) && isfield(M_.(model_type), model_name))
error(['Could not find M_.' model_type '.' model_name ...
'. First declare it via the var_model or ' ...
'trend_component_model statement.']);
end
end
%% Call Dynamic Function
[~, g1] = feval([M_.fname '.dynamic'], ...
ones(max(max(M_.lead_lag_incidence)), 1), ...
ones(1, M_.exo_nbr), ...
M_.params, ...
zeros(M_.endo_nbr, 1), ...
1);
% Choose rows of Jacobian based on equation tags
ntags = length(M_.(model_type).(model_name).eqtags);
g1rows = zeros(ntags, 1);
for i = 1:ntags
idxs = strcmp(M_.equations_tags(:, 3), M_.(model_type).(model_name).eqtags{i});
if any(idxs)
g1rows(i) = M_.equations_tags{idxs, 1};
end
end
g1 = -1 * g1(g1rows, :);
% Check for leads
if rows(M_.lead_lag_incidence) == 3
idxs = M_.lead_lag_incidence(3, M_.lead_lag_incidence(3, :) ~= 0);
assert(~any(any(g1(g1rows, idxs))), ...
['You cannot have leads in the equations specified by ' strjoin(M_.(model_type).(model_name).eqtags, ',')]);
end
%% Organize AR & EC matrices
assert(length(M_.(model_type).(model_name).lhs) == rows(g1));
% Find RHS vars for AR & EC matrices
ecRhsVars = [];
lhs = M_.(model_type).(model_name).lhs;
rhsvars = cell(length(lhs), 1);
for i = 1:length(M_.(model_type).(model_name).rhs.vars_at_eq)
vars = M_.(model_type).(model_name).rhs.vars_at_eq{i}.var;
rhsvars{i}.vars = vars;
rhsvars{i}.lags = M_.(model_type).(model_name).rhs.vars_at_eq{i}.lag;
rhsvars{i}.arRhsIdxs = [];
rhsvars{i}.ecRhsIdxs = [];
rhsvars{i}.ecRhsVars = [];
for j = 1:length(vars)
if vars(j) <= M_.orig_endo_nbr
% vars(j) is not an aux var
if ismember(vars(j), lhs)
rhsvars{i}.arRhsIdxs = [rhsvars{i}.arRhsIdxs find(lhs == vars(j))];
rhsvars{i}.ecRhsIdxs = [rhsvars{i}.ecRhsIdxs -1];
rhsvars{i}.ecRhsVars = [rhsvars{i}.ecRhsVars -1];
else
ecRhsVars = union(ecRhsVars, vars(j));
rhsvars{i}.ecRhsVars = [rhsvars{i}.ecRhsVars vars(j)];
rhsvars{i}.arRhsIdxs = [rhsvars{i}.arRhsIdxs -1];
rhsvars{i}.ecRhsIdxs = [rhsvars{i}.ecRhsIdxs find(rhsvars{i}.ecRhsVars == vars(j))];
end
else
% Search aux vars for matching lhs var
lhsvaridx = findLhsInAuxVar(vars(j), lhs);
if lhsvaridx >= 1
rhsvars{i}.arRhsIdxs = [rhsvars{i}.arRhsIdxs find(lhs == lhsvaridx)];
rhsvars{i}.ecRhsIdxs = [rhsvars{i}.ecRhsIdxs -1];
rhsvars{i}.ecRhsVars = [rhsvars{i}.ecRhsVars -1];
else
% otherwise find endog that corresponds to this aux var
varidx = findVarNoLag(vars(j));
ecRhsVars = union(ecRhsVars, varidx);
rhsvars{i}.ecRhsVars = [rhsvars{i}.ecRhsVars varidx];
rhsvars{i}.arRhsIdxs = [rhsvars{i}.arRhsIdxs -1];
rhsvars{i}.ecRhsIdxs = [rhsvars{i}.ecRhsIdxs find(rhsvars{i}.ecRhsVars == varidx)];
end
end
end
end
[rhsvars, ecRhsVars] = reorderECvars(rhsvars, ecRhsVars, lhs);
% Initialize matrices
oo_.(model_type).(model_name).ar = zeros(length(lhs), length(lhs), max(M_.(model_type).(model_name).max_lag));
oo_.(model_type).(model_name).ec = zeros(length(lhs), length(ecRhsVars), 1);
oo_.(model_type).(model_name).ar_idx = lhs;
oo_.(model_type).(model_name).ec_idx = ecRhsVars;
% Fill matrices
for i = 1:length(lhs)
for j = 1:length(rhsvars{i}.vars)
var = rhsvars{i}.vars(j);
if rhsvars{i}.lags(j) == -1
g1col = M_.lead_lag_incidence(1, var);
else
g1col = M_.lead_lag_incidence(2, var);
end
if g1col ~= 0 && any(g1(:, g1col))
if rhsvars{i}.arRhsIdxs(j) > 0
% Fill AR
lag = findLagForVar(var, -rhsvars{i}.lags(j), 0, lhs);
oo_.(model_type).(model_name).ar(i, rhsvars{i}.arRhsIdxs(j), lag) = ...
oo_.(model_type).(model_name).ar(i, rhsvars{i}.arRhsIdxs(j), lag) + g1(i, g1col);
elseif rhsvars{i}.ecRhsIdxs(j) > 0
% Fill EC
lag = findLagForVar(var, -rhsvars{i}.lags(j), 0, ecRhsVars);
if lag==1
if size(oo_.(model_type).(model_name).ec, 3) < lag
oo_.(model_type).(model_name).ec(i, rhsvars{i}.ecRhsIdxs(j), lag) = 0;
end
oo_.(model_type).(model_name).ec(i, rhsvars{i}.ecRhsIdxs(j), lag) = ...
oo_.(model_type).(model_name).ec(i, rhsvars{i}.ecRhsIdxs(j), lag) + g1(i, g1col);
end
else
error('Shouldn''t arrive here');
end
end
end
end
end
function [rhsvars, ecRhsVarsReordered] = reorderECvars(rhsvars, ecRhsVars, lhs)
global M_
ecRhsVarsReordered = [];
for i = 1:length(lhs)
av = M_.aux_vars([M_.aux_vars.endo_index] == lhs(i));
if ~isempty(av)
var = ecRhsVars(ecRhsVars == av.orig_index);
if ~isempty(var)
ecRhsVarsReordered = [ecRhsVarsReordered var];
for j = 1:length(rhsvars)
rhsidx = find(rhsvars{j}.ecRhsVars == var);
if ~isempty(rhsidx)
rhsvars{j}.ecRhsIdxs(rhsidx) = length(ecRhsVarsReordered);
end
end
end
end
end
end
function lhsvaridx = findLhsInAuxVar(auxVar, lhsvars)
global M_
if auxVar <= M_.orig_endo_nbr
lhsvaridx = -1;
return
end
av = M_.aux_vars([M_.aux_vars.endo_index] == auxVar);
if av.type == 8 || av.type == 10
if ismember(av.endo_index, lhsvars)
lhsvaridx = av.endo_index;
else
lhsvaridx = findLhsInAuxVar(av.orig_index, lhsvars);
end
else
if ismember(av.orig_index, lhsvars)
lhsvaridx = av.orig_index;
else
lhsvaridx = findLhsInAuxVar(av.orig_index, lhsvars);
end
end
end
function idx = findVarNoLag(auxVar)
global M_
if auxVar <= M_.orig_endo_nbr
error('Shouldn''t arrive here')
end
av = M_.aux_vars([M_.aux_vars.endo_index] == auxVar);
if av.type == 8 || av.type == 10
idx = av.endo_index;
else
if av.orig_index <= M_.orig_endo_nbr
idx = av.orig_index;
else
idx = findVarNoLag(av.orig_index);
end
end
end
function [lag, ndiffs] = findLagForVar(auxVar, lag, ndiffs, rhsVars)
global M_
if auxVar <= M_.orig_endo_nbr
assert(lag > 0)
return
end
av = M_.aux_vars([M_.aux_vars.endo_index] == auxVar);
if av.type == 8
ndiffs = ndiffs + 1;
end
if ismember(av.endo_index, rhsVars)
if av.type == 8 || av.type == 9
lag = lag + abs(av.orig_lead_lag);
end
elseif ismember(av.orig_index, rhsVars)
if av.orig_index <= M_.orig_endo_nbr
lag = lag + abs(av.orig_lead_lag);
else
[lag, ndiffs] = findLagForVar(av.orig_index, lag + max(1, abs(av.orig_lead_lag)), ndiffs, rhsVars);
end
else
if av.type == 8
[lag, ndiffs] = findLagForVar(av.orig_index, lag, ndiffs, rhsVars);
else
[lag, ndiffs] = findLagForVar(av.orig_index, lag + max(1, abs(av.orig_lead_lag)), ndiffs, rhsVars);
end
end
assert(lag > 0)
end
|