stepan
/
dynare
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

OSR: Remove globals and wrapper layer, move to +subfolder

kalman-mex
Johannes Pfeifer 2023-09-08 11:55:15 +02:00 committed by Sébastien Villemot
parent d601ca4a2e
commit 2b240210d0
No known key found for this signature in database
GPG Key ID: 2CECE9350ECEBE4A
5 changed files with 85 additions and 136 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

30
matlab/get_variance_of_endogenous_variables.m → matlab/+osr/get_variance_of_endogenous_variables.m
View File

@ -1,19 +1,21 @@
function vx1 = get_variance_of_endogenous_variables(dr,i_var)
% function vx1 = get_variance_of_endogenous_variables(dr,i_var)
function vx1 = get_variance_of_endogenous_variables(M_,options_,dr,i_var)
% vx1 = get_variance_of_endogenous_variables(dr,i_var)
% Gets the variance of a variables subset
%
% INPUTS
% dr: structure of decisions rules for stochastic simulations
% i_var: indices of a variables list
% M_ [structure] Dynare's model structure
% oo_ [structure] Dynare's results structure
% options_ [structure] Dynare's options structure
% dr: [structure] structure of decisions rules for stochastic simulations
% i_var: [integer] indices of a variables list
%
% OUTPUTS
% vx1: variance-covariance matrix
% vx1: [double] variance-covariance matrix
%
% SPECIAL REQUIREMENTS
% none
% Copyright © 2003-2017 Dynare Team
% Copyright © 2003-2023 Dynare Team
%
% This file is part of Dynare.
%
@ -30,22 +32,14 @@ function vx1 = get_variance_of_endogenous_variables(dr,i_var)
% 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_ options_
endo_nbr = M_.endo_nbr;
Sigma_e = M_.Sigma_e;
nstatic = M_.nstatic;
nspred = M_.nspred;
ghx = dr.ghx(i_var,:);
ghu = dr.ghu(i_var,:);
nc = size(ghx,2);
n = length(i_var);
[A,B] = kalman_transition_matrix(dr,nstatic+(1:nspred),1:nc);
[A,B] = kalman_transition_matrix(dr,M_.nstatic+(1:M_.nspred),1:nc);
[vx,u] = lyapunov_symm(A,B*Sigma_e*B',options_.lyapunov_fixed_point_tol,options_.qz_criterium,options_.lyapunov_complex_threshold, [], options_.debug);
[vx,u] = lyapunov_symm(A,B*M_.Sigma_e*B',options_.lyapunov_fixed_point_tol,options_.qz_criterium,options_.lyapunov_complex_threshold, [], options_.debug);
if size(u,2) > 0
i_stat = find(any(abs(ghx*u) < options_.schur_vec_tol,2)); %only set those variances of objective function for which variance is finite
@ -56,4 +50,4 @@ else
end
vx1 = Inf*ones(n,n);
vx1(i_stat,i_stat) = ghx*vx*ghx'+ghu*Sigma_e*ghu';
vx1(i_stat,i_stat) = ghx*vx*ghx'+ghu*M_.Sigma_e*ghu';

19
matlab/osr_obj.m → matlab/+osr/objective.m
View File

@ -1,8 +1,11 @@
function [loss,info,exit_flag,vx,junk]=osr_obj(x,i_params,i_var,weights)
function [loss,info,exit_flag,vx,junk]=objective(x,M_, oo_, options_,i_params,i_var,weights)
% objective function for optimal simple rules (OSR)
% INPUTS
% x vector values of the parameters
% over which to optimize
% M_ [structure] Dynare's model structure
% oo_ [structure] Dynare's results structure
% options_ [structure] Dynare's options structure
% i_params vector index of optimizing parameters in M_.params
% i_var vector variables indices
% weights vector weights in the OSRs
@ -17,7 +20,7 @@ function [loss,info,exit_flag,vx,junk]=osr_obj(x,i_params,i_var,weights)
%
% SPECIAL REQUIREMENTS
% none
% Copyright © 2005-2017 Dynare Team
% Copyright © 2005-2023 Dynare Team
%
% This file is part of Dynare.
%
@ -34,20 +37,14 @@ function [loss,info,exit_flag,vx,junk]=osr_obj(x,i_params,i_var,weights)
% 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_ options_ optimal_Q_ it_
% global ys_ Sigma_e_ endo_nbr exo_nbr optimal_Q_ it_ ykmin_ options_
junk = [];
exit_flag = 1;
vx = [];
info=zeros(4,1);
loss=[];
% set parameters of the policiy rule
M_.params(i_params) = x;
% don't change below until the part where the loss function is computed
it_ = M_.maximum_lag+1;
[dr,info,M_,oo_] = resol(0,M_,options_,oo_);
[dr,info] = resol(0,M_,options_,oo_);
if info(1)
if info(1) == 3 || info(1) == 4 || info(1) == 5 || info(1)==6 ||info(1) == 19 ||...
@ -63,5 +60,5 @@ if info(1)
end
end
vx = get_variance_of_endogenous_variables(dr,i_var);
loss = full(weights(:)'*vx(:));
vx = osr.get_variance_of_endogenous_variables(M_,options_,dr,i_var);
loss = full(weights(:)'*vx(:));

105
matlab/osr1.m → matlab/+osr/run.m
View File

@ -1,22 +1,32 @@
function osr_res = osr1(i_params,i_var,weights)
% Compute the Optimal Simple Rules
function [info, oo_, options_, M_] = run(M_, options_, oo_, var_list, params, i_var,W)
% [info, oo_, options_, M_] = osr(M_, options_, oo_, var_list, params, i_var,W)
% Function computing the solution to the optimal simple rule-problem
%
% INPUTS
% i_params vector index of optimizing parameters in M_.params
% i_var vector variables indices in declaration order
% weights vector weights in the OSRs
% M_ [structure] Dynare's model structure
% oo_ [structure] Dynare's results structure
% options_ [structure] Dynare's options structure
% var_list [character array] list of endogenous variables specified
% params [character array] list of parameter to be chosen in
% optimal simple rule
% i_var [n_osr_vars by 1 double] indices of osr-variable in
% specified in optim_weights in declaration order
% W [M_.endo_nbr by M_.endo_nbr sparse matrix] Weighting matrix for variance of endogenous variables
%
% OUTPUTS
% osr_res: [structure] results structure containing:
% - objective_function [scalar double] value of the objective
% function at the optimum
% - optim_params [structure] parameter values at the optimum
% info [integer] scalar or vector, error code.
% oo_ [structure] Dynare's results structure, containing subfield
% osr_res: results structure containing:
% - objective_function [scalar double] value of the objective
% function at the optimum
% - optim_params [structure] parameter values at the optimum
% options_ [structure] Dynare's options structure
% M_ [structure] Dynare's model structure
%
% Algorithm:
%
% Uses Newton-type optimizer csminwel to directly solve quadratic
% osr-problem
%
% Copyright © 2005-2019 Dynare Team
% SPECIAL REQUIREMENTS
% none.
% Copyright © 2001-2023 Dynare Team
%
% This file is part of Dynare.
%
@ -33,9 +43,24 @@ function osr_res = osr1(i_params,i_var,weights)
% 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_ options_ it_
options_.order = 1;
it_ = M_.maximum_lag + 1 ;
if isempty(options_.qz_criterium)
options_.qz_criterium = 1+1e-6;
end
oo_=make_ex_(M_,options_,oo_);
np = size(params,1);
i_params = zeros(np,1);
for i=1:np
str = deblank(params(i,:));
i_params(i) = strmatch(str{:}, M_.param_names, 'exact');
end
if ~options_.noprint
fprintf('\nOPTIMAL SIMPLE RULE\n')
end
osr_res.error_indicator = 1; %initialize indicator
@ -43,7 +68,7 @@ if M_.exo_nbr == 0
oo_.exo_steady_state = [] ;
end
if ~ M_.lead_lag_incidence(M_.maximum_lag+1,:) > 0
if ~M_.lead_lag_incidence(M_.maximum_lag+1,:) > 0
error ('OSR: Error in model specification: some variables don''t appear as current') ;
end
@ -51,7 +76,7 @@ if M_.maximum_lead == 0
error ('OSR: Backward or static model: no point in using OSR') ;
end
if any(any(isinf(weights)))
if any(any(isinf(W)))
error ('OSR: At least one of the optim_weights is infinite.') ;
end
@ -81,19 +106,13 @@ if isfield(options_.osr,'maxit') || isfield(options_.osr,'tolf')
end
oo_.dr = set_state_space(oo_.dr,M_,options_);
np = size(i_params,1);
t0 = M_.params(i_params);
par_0 = M_.params(i_params);
inv_order_var = oo_.dr.inv_order_var;
%extract unique entries of covariance
i_var=unique(i_var);
%% do initial checks
[loss,info]=osr_obj(t0,i_params,inv_order_var(i_var),weights(i_var,i_var));
[loss,info]=osr.objective(par_0,M_,oo_,options_,i_params,inv_order_var(i_var),W(i_var,i_var));
if info~=0
print_info(info, options_.noprint, options_);
else
@ -108,7 +127,6 @@ if ~options_.noprint && isinf(loss)
error('OSR: Initial likelihood is infinite')
end
if isequal(options_.osr.opt_algo,5)
error('OSR: OSR does not support opt_algo=5.')
elseif isequal(options_.osr.opt_algo,6)
@ -118,13 +136,12 @@ elseif isequal(options_.osr.opt_algo,10)
elseif isequal(options_.osr.opt_algo,11)
error('OSR: OSR does not support opt_algo=11.')
else
if ~isempty(M_.osr.param_bounds) && ~(ismember(options_.osr.opt_algo,[1,2,5,9]) || ischar(options_.osr.opt_algo))
error('OSR: OSR with bounds on parameters requires a constrained optimizer, i.e. opt_algo= 1,2 or 9.')
end
%%do actual optimization
[p, f] = dynare_minimize_objective(str2func('osr_obj'),t0,options_.osr.opt_algo,options_,M_.osr.param_bounds,M_.param_names(i_params),[],[], i_params,...
inv_order_var(i_var),weights(i_var,i_var));
[p, f] = dynare_minimize_objective(str2func('osr.objective'),par_0,options_.osr.opt_algo,options_,M_.osr.param_bounds,M_.param_names(i_params),[],[], M_,oo_,options_,i_params,...
inv_order_var(i_var),W(i_var,i_var));
end
osr_res.objective_function = f;
@ -134,16 +151,22 @@ for i=1:length(i_params)
end
if ~options_.noprint
skipline()
disp('OPTIMAL VALUE OF THE PARAMETERS:')
skipline()
for i=1:np
fprintf('%16s %16.6g\n\n', M_.param_names{i_params(i)}, p(i));
my_title='OPTIMAL VALUE OF THE PARAMETERS';
labels = M_.param_names(i_params);
headers = {'Variables'; 'Value'};
lh = cellofchararraymaxlength(labels)+2;
dyntable(options_, my_title, headers, labels, p, lh, 10, 6);
if options_.TeX
labels = M_.param_names_tex(i_params);
lh = cellofchararraymaxlength(labels)+2;
dyn_latex_table(M_, options_, my_title, 'osr', headers, labels, p, lh, 10, 6);
end
fprintf('Objective function : %16.6g\n\n',f);
skipline()
fprintf('\nObjective function : %16.6g\n\n',f);
end
[oo_.dr,info,M_,oo_] = resol(0,M_,options_,oo_);
if ~info
[info, oo_, options_, M_] = stoch_simul(M_, options_, oo_, var_list);
if ~info(1)
osr_res.error_indicator=0;
end
end
oo_.osr = osr_res;

65
matlab/osr.m
View File

@ -1,65 +0,0 @@
function osr_res = osr(var_list, params, i_var,W)
% osr_res = osr(var_list,params,i_var,W)
% Wrapper function computing the solution to the optimal simple
% rule-problem; calls osr1 for actual computation
% INPUTS
% var_list [character array] list of endogenous variables specified
% after osr1-command (deprecated and not used anymore)
% params [character array] list of parameter to be chosen in
% optimal simple rule
% i_var [n_osr_vars by 1 double] indices of osr-variable in
% specified in optim_weights in declaration order
% W [M_.endo_nbr by M_.endo_nbr sparse matrix] Weighting matrix for variance of endogenous variables
%
% OUTPUTS
% osr_res: [structure] results structure containing:
% - objective_function [scalar double] value of the objective
% function at the optimum
% - optim_params [structure] parameter values at the optimum
%
%
% SPECIAL REQUIREMENTS
% none.
% Copyright © 2001-2018 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_ options_ oo_
options_.order = 1;
if isempty(options_.qz_criterium)
options_.qz_criterium = 1+1e-6;
end
oo_=make_ex_(M_,options_,oo_);
np = size(params,1);
i_params = zeros(np,1);
for i=1:np
str = deblank(params(i,:));
i_params(i) = strmatch(str{:}, M_.param_names, 'exact');
end
if ~options_.noprint
skipline()
disp('OPTIMAL SIMPLE RULE')
skipline()
end
osr_res = osr1(i_params,i_var,W);
[~, oo_, options_, M_] = stoch_simul(M_, options_, oo_, var_list);

2
preprocessor

@ -1 +1 @@
Subproject commit 140c91e91188ee701d4a185d1ad771c23999f491
Subproject commit bd0ba65a61c0d97e3f537194136e3cdad0f4f3b2