/usr/share/dynare/matlab/simultxdet.m is in dynare-common 4.4.1-1build1.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
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 | function [y_,int_width]=simultxdet(y0,ex,ex_det, iorder,var_list,M_,oo_,options_)
%function [y_,int_width]=simultxdet(y0,ex,ex_det, iorder,var_list,M_,oo_,options_)
%
% Simulates a stochastic model in the presence of deterministic exogenous shocks
%
% INPUTS:
% y0: initial values, of length M_.maximum_lag
% ex: matrix of stochastic exogenous shocks, starting at period 1
% ex_det: matrix of deterministic exogenous shocks, starting at period 1-M_.maximum_lag
% iorder: order of approximation
% var_list: list of endogenous variables to simulate
%
% The forecast horizon is equal to size(ex, 1).
% The condition size(ex,1)+M_.maximum_lag=size(ex_det,1) must be verified
% for consistency.
% Copyright (C) 2008-2012 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/>.
dr = oo_.dr;
ykmin = M_.maximum_lag;
endo_nbr = M_.endo_nbr;
exo_det_steady_state = oo_.exo_det_steady_state;
nstatic = M_.nstatic;
nspred = M_.nspred;
nc = size(dr.ghx,2);
iter = size(ex,1);
if size(ex_det, 1) ~= iter+ykmin
error('Size mismatch: number of forecasting periods for stochastic exogenous and deterministic exogenous don''t match')
end
nx = size(dr.ghu,2);
y_ = zeros(size(y0,1),iter+ykmin);
y_(:,1:ykmin) = y0;
k1 = [ykmin:-1:1];
k2 = dr.kstate(find(dr.kstate(:,2) <= ykmin+1),[1 2]);
k2 = k2(:,1)+(ykmin+1-k2(:,2))*endo_nbr;
k3 = M_.lead_lag_incidence(1:ykmin,:)';
k3 = find(k3(:));
k4 = dr.kstate(find(dr.kstate(:,2) < ykmin+1),[1 2]);
k4 = k4(:,1)+(ykmin+1-k4(:,2))*endo_nbr;
nvar = size(var_list,1);
if nvar == 0
nvar = endo_nbr;
ivar = [1:nvar];
else
ivar=zeros(nvar,1);
for i=1:nvar
i_tmp = strmatch(var_list(i,:),M_.endo_names,'exact');
if isempty(i_tmp)
disp(var_list(i,:));
error (['One of the variable specified does not exist']) ;
else
ivar(i) = i_tmp;
end
end
end
if iorder == 1
for i = ykmin+1: iter+ykmin
tempx1 = y_(dr.order_var,k1);
tempx2 = tempx1-repmat(dr.ys(dr.order_var),1,ykmin);
tempx = tempx2(k2);
y_(dr.order_var,i) = dr.ys(dr.order_var)+dr.ghx*tempx+dr.ghu* ...
ex(i-ykmin,:)';
for j=1:min(ykmin+M_.exo_det_length+1-i,M_.exo_det_length)
y_(dr.order_var,i) = y_(dr.order_var,i) + dr.ghud{j}*(ex_det(i+j-1,:)'-exo_det_steady_state);
end
k1 = k1+1;
end
elseif iorder == 2
for i = ykmin+1: iter+ykmin
tempx1 = y_(dr.order_var,k1);
tempx2 = tempx1-repmat(dr.ys(dr.order_var),1,ykmin);
tempx = tempx2(k2);
tempu = ex(i-ykmin,:)';
tempuu = kron(tempu,tempu);
tempxx = kron(tempx,tempx);
tempxu = kron(tempx,tempu);
y_(dr.order_var,i) = dr.ys(dr.order_var)+dr.ghs2/2+dr.ghx*tempx+ ...
dr.ghu*tempu+0.5*(dr.ghxx*tempxx+dr.ghuu*tempuu)+dr.ghxu* ...
tempxu;
for j=1:min(ykmin+M_.exo_det_length+1-i,M_.exo_det_length)
tempud = ex_det(i+j-1,:)'-exo_det_steady_state;
tempudud = kron(tempud,tempud);
tempxud = kron(tempx,tempud);
tempuud = kron(tempu,tempud);
y_(dr.order_var,i) = y_(dr.order_var,i) + dr.ghud{j}*tempud + ...
dr.ghxud{j}*tempxud + dr.ghuud{j}*tempuud + ...
0.5*dr.ghudud{j,j}*tempudud;
for k=1:j-1
tempudk = ex_det(i+k-1,:)'-exo_det_steady_state;
tempududk = kron(tempudk,tempud);
y_(dr.order_var,i) = y_(dr.order_var,i) + ...
dr.ghudud{k,j}*tempududk;
end
end
k1 = k1+1;
end
end
[A,B] = kalman_transition_matrix(dr,nstatic+(1:nspred),1:nc,M_.exo_nbr);
inv_order_var = dr.inv_order_var;
ghx1 = dr.ghx(inv_order_var(ivar),:);
ghu1 = dr.ghu(inv_order_var(ivar),:);
sigma_u = B*M_.Sigma_e*B';
sigma_u1 = ghu1*M_.Sigma_e*ghu1';
sigma_y = 0;
for i=1:iter
sigma_y1 = ghx1*sigma_y*ghx1'+sigma_u1;
var_yf(i,:) = diag(sigma_y1)';
if i == iter
break
end
sigma_u = A*sigma_u*A';
sigma_y = sigma_y+sigma_u;
end
fact = norminv((1-options_.conf_sig)/2,0,1);
int_width = zeros(iter,endo_nbr);
for i=1:nvar
int_width(:,i) = fact*sqrt(var_yf(:,i));
end
|