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

Merge branch 'master' into ecb-master 

Fixed conflict in dseries submodule.
time-shift
Stéphane Adjemian (Charybdis) 2018-02-10 12:53:34 +01:00
parent 42417d992d 350b8fb529
commit 3a052f500c
42 changed files with 462 additions and 262 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

61
NEWS
View File

@ -1,3 +1,56 @@
Announcement for Dynare 4.5.4 (on 2018-01-29)
=============================================
We are pleased to announce the release of Dynare 4.5.4.
This is a bugfix release.
The Windows packages are already available for download at:
http://www.dynare.org/download/dynare-stable
The Mac and GNU/Linux packages (for Debian and Ubuntu LTS) should follow soon.
This release is compatible with MATLAB versions 7.5 (R2007b) to 9.3 (R2017b)
and with GNU Octave versions 4.2.
Here is a list of the problems identified in version 4.5.3 and that have been
fixed in version 4.5.4:
- The `type` option of `plot_shock_decomposition` was always set to `qoq` regardless of what is specified.
- Bug in GSA when no parameter was detected below pvalue threshold.
- Various bug fixes in shock decompositions.
- Bug in reading in macro arrays passed on `dynare` command line via the `-D` option.
- Estimation with missing values was crashing if the `prefilter` option was used.
- Added a workaround for a difference in behaviour between Octave and Matlab regarding the creation
of function handles for functions that do not exist in the path. With Octave 4.2.1, steady state
files did not work if no auxiliary variables were created.
- The `stoch_simul` command was crashing with a cryptic message if option `order=3` was used without
setting `k_order_solver`.
- In cases where the prior bounds are infinite and the mode is estimated at exactly 0, no `mode_check`
graphs were displayed.
- Parallel execution of MCMC was broken in models without auxiliary variables.
- Reading data with column names from Excel might crash.
- The multivariate Kalman smoother was crashing in case of missing data in the observations and
`Finf` became singular.
- The `plot_shock_decomposition` command ignored various user-defined options like `fig_name`,
`use_shock_groups` or `interactive` and instead used the default options.
- Nested `@#ifdef` and `@#ifndef` statements don't work in the macroprocessor.
Announcement for Dynare 4.5.3 (on 2017-10-19)
=============================================
@ -12,7 +65,7 @@ The Windows packages are already available for download at:
The Mac and GNU/Linux packages (for Debian and Ubuntu LTS) should follow soon.
This release is compatible with MATLAB versions 7.3 (R2006b) to 9.3 (R2017b)
This release is compatible with MATLAB versions 7.5 (R2007b) to 9.3 (R2017b)
and with GNU Octave versions 4.2.
Here is a list of the problems identified in version 4.5.2 and that have been
@ -37,7 +90,7 @@ The Windows packages are already available for download at:
The Mac and GNU/Linux packages (for Debian and Ubuntu LTS) should follow soon.
This release is compatible with MATLAB versions 7.3 (R2006b) to 9.3 (R2017b)
This release is compatible with MATLAB versions 7.5 (R2007b) to 9.3 (R2017b)
and with GNU Octave versions 4.2.
Here is a list of the problems identified in version 4.5.1 and that have been
@ -116,7 +169,7 @@ The Windows packages are already available for download at:
The Mac and GNU/Linux packages (for Debian and Ubuntu LTS) should follow soon.
This release is compatible with MATLAB versions 7.3 (R2006b) to 9.2 (R2017a)
This release is compatible with MATLAB versions 7.5 (R2007b) to 9.2 (R2017a)
and with GNU Octave versions 4.2.
Here is a list of the problems identified in version 4.5.0 and that have been
@ -162,7 +215,7 @@ The Mac and Debian/Ubuntu packages should follow soon.
All users are strongly encouraged to upgrade.
This release is compatible with MATLAB versions ranging from 7.3 (R2006b) to
This release is compatible with MATLAB versions ranging from 7.5 (R2007b) to
9.2 (R2017a) and with GNU Octave version 4.2.
Here is the list of major user-visible changes:

11
doc/dynare.texi
View File

@ -7602,9 +7602,9 @@ model).
(@xref{plot_shock_decomposition}).
@item presample = @var{INTEGER}
@anchor{presample_shock_decomposition} First data point from which recursive
@anchor{presample_shock_decomposition} Data point above which recursive
realtime shock decompositions are computed, @i{i.e.} for
@math{T=[@code{presample}@dots{}@code{nobs}]}.
@math{T=[@code{presample+1}@dots{}@code{nobs}]}.
@item forecast = @var{INTEGER}
@anchor{forecast_shock_decomposition} Compute shock decompositions up to
@ -7613,6 +7613,13 @@ realtime shock decompositions are computed, @i{i.e.} for
@item save_realtime = @var{INTEGER_VECTOR}
@anchor{save_realtime} Choose for which vintages to save the full realtime
shock decomposition. Default: @math{0}.
@item fast_realtime = @var{INTEGER}
@anchor{fast_realtime} Runs the smoother only twice: once for the last in-sample
and once for the last out-of-sample data point, where the provided integer defines the last observation
(equivalent to @ref{nobs}).
Default: not enabled.
@end table
@outputhead

4
matlab/DsgeSmoother.m
View File

@ -58,7 +58,7 @@ function [alphahat,etahat,epsilonhat,ahat,SteadyState,trend_coeff,aK,T,R,P,PK,de
% SPECIAL REQUIREMENTS
% None
% Copyright (C) 2006-2017 Dynare Team
% Copyright (C) 2006-2018 Dynare Team
%
% This file is part of Dynare.
%
@ -183,7 +183,7 @@ elseif options_.lik_init == 3 % Diffuse Kalman filter
Z = [Z, eye(vobs)];
end
end
[Pstar,Pinf] = compute_Pinf_Pstar(mf,T,R,Q,options_.qz_criterium,oo_.dr.restrict_var_list);
[Pstar,Pinf] = compute_Pinf_Pstar(mf,T,R,Q,options_.qz_criterium);
elseif options_.lik_init == 4 % Start from the solution of the Riccati equation.
[err, Pstar] = kalman_steady_state(transpose(T),R*Q*transpose(R),transpose(build_selection_matrix(mf,np,vobs)),H);
mexErrCheck('kalman_steady_state',err);

7
matlab/check_list_of_variables.m
View File

@ -31,8 +31,7 @@ function varlist = check_list_of_variables(options_, M_, varlist)
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
%get uniques
% Get uniques
[junk1, junk2, index_uniques] = varlist_indices(varlist, M_.endo_names);
varlist = varlist(index_uniques);
@ -130,7 +129,7 @@ elseif isempty(varlist) && isempty(options_.endo_vars_for_moment_computations_in
elseif choice==2
varlist = options_.varobs;
elseif choice==3
varlist = NaN;
varlist = cell(0);
else
skipline()
disp('YOU HAVE TO ANSWER 1, 2 or 3!')
@ -138,7 +137,7 @@ elseif isempty(varlist) && isempty(options_.endo_vars_for_moment_computations_in
end
end
end
if isnan(varlist)
if isempty(varlist)
edit([M_.fname '.mod'])
end
skipline()

76
matlab/compute_Pinf_Pstar.m
View File

@ -30,7 +30,7 @@ function [Pstar,Pinf] = compute_Pinf_Pstar(mf,T,R,Q,qz_criterium, restrict_colum
% SPECIAL REQUIREMENTS
% None
% Copyright (C) 2006-2017 Dynare Team
% Copyright (C) 2006-2018 Dynare Team
%
% This file is part of Dynare.
%
@ -48,8 +48,22 @@ function [Pstar,Pinf] = compute_Pinf_Pstar(mf,T,R,Q,qz_criterium, restrict_colum
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
np = size(T,1);
if nargin == 6
indx = restrict_columns;
indx0=find(~ismember([1:np],indx));
else
indx=(find(max(abs(T))>=1.e-10));
indx0=(find(max(abs(T))<1.e-10));
end
np0=length(indx0);
Tbkp = T;
T0=T(indx0,indx); % static variables vs. dynamic ones
R0=R(indx0,:); % matrix of shocks for static variables
% perform Kitagawa transformation
% Perform Kitagawa transformation only for non-zero columns of T
T=T(indx,indx);
R=R(indx,:);
np = size(T,1);
[QT,ST] = schur(T);
e1 = abs(ordeig(ST)) > 2-qz_criterium;
[QT,ST] = ordschur(QT,ST,e1);
@ -59,7 +73,6 @@ nk1 = nk+1;
Pstar = zeros(np,np);
R1 = QT'*R;
B = R1*Q*R1';
% computes variance of stationary block (lower right)
i = np;
while i >= nk+2
if ST(i,i-1) == 0
@ -100,13 +113,64 @@ if i == nk+1
Pstar(nk1,nk1)=(B(nk1,nk1)+c)/(1-ST(nk1,nk1)*ST(nk1,nk1));
end
if np0
ST1=ST;
% Now I recover stationarized static variables using
% ss = s-A*z
% and
% z-z(-1) (growth rates of unit roots) only depends on stationary variables
Pstar = blkdiag(zeros(np0),Pstar);
ST = [zeros(length(Pstar),length(indx0)) [T0*QT ;ST]];
R1 = [R0; R1];
% Build the matrix for stationarized variables
STinf = ST(np0+1:np0+nk,np0+1:np0+nk);
iSTinf = inv(STinf);
ST0=ST;
ST0(:,1:np0+nk)=0; % stationarized static + 1st difference only respond to lagged stationary states
ST00 = ST(1:np0,np0+1:np0+nk);
% A\B is the matrix division of A into B, which is roughly the
% same as INV(A)*B
ST0(1:np0,np0+nk+1:end) = ST(1:np0,np0+nk+1:end)-ST00*(iSTinf*ST(np0+1:np0+nk,np0+nk+1:end)); % snip non-stationary part
R10 = R1;
R10(1:np0,:) = R1(1:np0,:)-ST00*(iSTinf*R1(np0+1:np0+nk,:)); % snip non-stationary part
% Kill non-stationary part before projecting Pstar
ST0(np0+1:np0+nk,:)=0;
R10(np0+1:np0+nk,:)=0; % is this questionable???? IT HAS TO in order to match Michel's version!!!
% project Pstar onto static x
Pstar = ST0*Pstar*ST0'+R10*Q*R10';
% QT(1:np0,np0+1:np0+nk) = QT(1:np0,np0+1:np0+nk)+ST(1:np0,np0+1:np0+nk); %%% is this questionable ????
% reorder QT entries
else
STinf = ST(np0+1:np0+nk,np0+1:np0+nk);
end
% stochastic trends with no influence on observed variables are
% arbitrarily initialized to zero
Pinf = zeros(np,np);
Pinf(1:nk,1:nk) = eye(nk);
for k = 1:nk
if norm(QT(mf,:)*ST(:,k)) < 1e-8
Pinf(k,k) = 0;
if np0
STtriu = STinf-eye(nk);
% A\B is the matrix division of A into B, which is roughly the
% same as INV(A)*B
STinf0 = ST00*(eye(nk)-iSTinf*STtriu);
Pinf = blkdiag(zeros(np0),Pinf);
QT = blkdiag(eye(np0),QT);
QTinf = QT;
QTinf(1:np0,np0+1:np0+nk) = STinf0;
QTinf([indx0(:); indx(:)],:) = QTinf;
STinf1 = [zeros(np0+np,np0) [STinf0; eye(nk); zeros(np-nk,nk)] zeros(np0+np,np-nk)];
for k = 1:nk
if norm(QTinf(mf,:)*ST([indx0(:); indx(:)],k+np0)) < 1e-8
Pinf(k+np0,k+np0) = 0;
end
end
Pinf = STinf1*Pinf*STinf1';
QT([indx0(:); indx(:)],:) = QT;
else
for k = 1:nk
if norm(QT(mf,:)*ST(:,k)) < 1e-8
Pinf(k+np0,k+np0) = 0;
end
end
end

6
matlab/discretionary_policy_engine.m
View File

@ -48,7 +48,7 @@ function [H,G,retcode]=discretionary_policy_engine(AAlag,AA0,AAlead,BB,bigw,inst
% Dennis, Richard (2007): Optimal policy in rational expectations models: new solution algorithms,
% Macroeconomic Dynamics, 11, 31–55.
% Copyright (C) 2007-2017 Dynare Team
% Copyright (C) 2007-2018 Dynare Team
%
% This file is part of Dynare.
%
@ -112,9 +112,9 @@ F1=F10;
while 1
iter=iter+1;
P=SylvesterDoubling(W+beta*F1'*Q*F1,beta*H1',H1,discretion_tol,solve_maxit);
if any(any(isnan(P)))
if any(any(isnan(P))) || any(any(isinf(P)))
P=SylvesterHessenbergSchur(W+beta*F1'*Q*F1,beta*H1',H1);
if any(any(isnan(P)))
if any(any(isnan(P))) || any(any(isinf(P)))
retcode=2;
return
end

4
matlab/disp_steady_state.m
View File

@ -32,9 +32,9 @@ function disp_steady_state(M,oo)
skipline()
disp('STEADY-STATE RESULTS:')
skipline()
endo_names = M.endo_names;
endo_names = char(M.endo_names);
steady_state = oo.steady_state;
for i = 1:M.orig_endo_nbr
disp(sprintf('%s \t\t %g', endo_names{i}, steady_state(i)));
fprintf('%s \t\t %g\n', endo_names(i,:), steady_state(i));
end

4
matlab/dsge_likelihood.m
View File

@ -115,7 +115,7 @@ function [fval,info,exit_flag,DLIK,Hess,SteadyState,trend_coeff,Model,DynareOpti
%! @end deftypefn
%@eod:
% Copyright (C) 2004-2017 Dynare Team
% Copyright (C) 2004-2018 Dynare Team
%
% This file is part of Dynare.
%
@ -372,7 +372,7 @@ switch DynareOptions.lik_init
error(['The model requires Diffuse filter, but you specified a different Kalman filter. You must set options_.kalman_algo ' ...
'to 0 (default), 3 or 4'])
end
[Pstar,Pinf] = compute_Pinf_Pstar(Z,T,R,Q,DynareOptions.qz_criterium,[1:length(T)]);
[Pstar,Pinf] = compute_Pinf_Pstar(Z,T,R,Q,DynareOptions.qz_criterium);
Z =zeros(length(BayesInfo.mf),size(T,1));
for i = 1:length(BayesInfo.mf)
Z(i,BayesInfo.mf(i))=1;

1
matlab/dynare_estimation_1.m
View File

@ -422,6 +422,7 @@ end
if (any(bayestopt_.pshape >0 ) && options_.mh_replic) || ...
(any(bayestopt_.pshape >0 ) && options_.load_mh_file) %% not ML estimation
bounds = prior_bounds(bayestopt_, options_.prior_trunc); %reset bounds as lb and ub must only be operational during mode-finding
outside_bound_pars=find(xparam1 < bounds.lb | xparam1 > bounds.ub);
if ~isempty(outside_bound_pars)
for ii=1:length(outside_bound_pars)

9
matlab/evaluate_steady_state_file.m
View File

@ -70,7 +70,14 @@ else
updated_params_flag = 0;
end
h_set_auxiliary_variables = str2func([M.fname '_set_auxiliary_variables']);
if M.set_auxiliary_variables
% Define function handle for the function setting the auxiliary
% variables only if the model has auxiliary variables. Otherwise
% Octave may crash (see https://savannah.gnu.org/bugs/?52568) because
% the function does not exist in the path.
h_set_auxiliary_variables = str2func([M.fname '_set_auxiliary_variables']);
end
if isnan(updated_params_flag) || (updated_params_flag && any(isnan(params(~isnan(params))-params1(~isnan(params))))) %checks if new NaNs were added
info(1) = 24;
info(2) = NaN;

15
matlab/expand_group.m
View File

@ -34,16 +34,19 @@ mydata=get(findobj(gcf,'tag',['group' int2str(ic)]),'userdata');
if isfield(mydata,'shock_decomp')
options.shock_decomp=mydata.shock_decomp;
end
options.plot_shock_decomp=mydata.plot_shock_decomp;
options.first_obs=mydata.first_obs;
options.nobs=mydata.nobs;
% define expanded group
label = mydata.shock_group.label;
shocks = mydata.shock_group.shocks;
options.shock_decomp.fig_name = [mydata.fig_name '. Expand'];
options.use_shock_groups = strrep(label,' ','_'); %[use_shock_groups_old int2str(ic)];
options.plot_shock_decomp.fig_name = [mydata.fig_name '. Expand'];
options.plot_shock_decomp.use_shock_groups = strrep(label,' ','_'); %[use_shock_groups_old int2str(ic)];
for j=1:length(shocks)
M.shock_groups.(options.use_shock_groups).(['group' int2str(j)]).label=shocks{j};
M.shock_groups.(options.use_shock_groups).(['group' int2str(j)]).shocks=shocks(j);
M.shock_groups.(options.plot_shock_decomp.use_shock_groups).(['group' int2str(j)]).label=shocks{j};
M.shock_groups.(options.plot_shock_decomp.use_shock_groups).(['group' int2str(j)]).shocks=shocks(j);
end
options.shock_decomp.interactive=0;
options.shock_decomp.expand=1;
options.plot_shock_decomp.interactive=0;
options.plot_shock_decomp.expand=1;
plot_shock_decomposition(M,oo,options,var_list_);

9
matlab/graph_decomp.m
View File

@ -193,12 +193,15 @@ for j=1:nvar
mydata.fig_name = DynareOptions.plot_shock_decomp.fig_name(2:end);
mydata.use_shock_groups = DynareOptions.plot_shock_decomp.use_shock_groups;
mydata.shock_group = shock_groups.(shock_ind{i});
mydata.shock_decomp = DynareOptions.plot_shock_decomp;
if ~isempty(mydata.shock_group.shocks{1})
mydata.shock_decomp = DynareOptions.shock_decomp;
mydata.plot_shock_decomp = DynareOptions.plot_shock_decomp;
mydata.first_obs = DynareOptions.first_obs;
mydata.nobs = DynareOptions.nobs;
if ~isempty(mydata.shock_group.shocks)
c = uicontextmenu;
hl.UIContextMenu=c;
browse_menu = uimenu(c,'Label','Browse group');
expand_menu = uimenu(c,'Label','Expand group','Callback',['expand_group(''' mydata.use_shock_groups ''',''' deblank(mydata.shock_decomp.orig_varlist(j,:)) ''',' int2str(i) ')']);
expand_menu = uimenu(c,'Label','Expand group','Callback',['expand_group(''' mydata.plot_shock_decomp.use_shock_groups ''',''' deblank(mydata.plot_shock_decomp.orig_varlist{j}) ''',' int2str(i) ')']);
set(expand_menu,'UserData',mydata,'Tag',['group' int2str(i)]);
for jmember = mydata.shock_group.shocks
uimenu('parent',browse_menu,'Label',char(jmember))

7
matlab/graph_decomp_detail.m
View File

@ -195,12 +195,15 @@ for j=1:nvar
mydata.fig_name = DynareOptions.plot_shock_decomp.fig_name(2:end);
mydata.use_shock_groups = DynareOptions.plot_shock_decomp.use_shock_groups;
mydata.shock_group = shock_groups.(shock_ind{ic});
mydata.shock_decomp = DynareOptions.plot_shock_decomp;
mydata.shock_decomp = DynareOptions.shock_decomp;
mydata.plot_shock_decomp = DynareOptions.plot_shock_decomp;
mydata.first_obs = DynareOptions.first_obs;
mydata.nobs = DynareOptions.nobs;
if ~isempty(mydata.shock_group.shocks)
c = uicontextmenu;
hax.UIContextMenu=c;
browse_menu = uimenu(c,'Label','Browse group');
expand_menu = uimenu(c,'Label','Expand group','Callback',['expand_group(''' mydata.use_shock_groups ''',''' deblank(mydata.shock_decomp.orig_varlist(j,:)) ''',' int2str(ic) ')']);
expand_menu = uimenu(c,'Label','Expand group','Callback',['expand_group(''' mydata.plot_shock_decomp.use_shock_groups ''',''' deblank(mydata.plot_shock_decomp.orig_varlist{j}) ''',' int2str(ic) ')']);
set(expand_menu,'UserData',mydata,'Tag',['group' int2str(ic)]);
for jmember = mydata.shock_group.shocks
uimenu('parent',browse_menu,'Label',char(jmember))

43
matlab/gsa/redform_map.m
View File

@ -232,10 +232,27 @@ for j = 1:length(anamendo)
istable=[1:length(iy)];
save([xdir,filesep, fname_ '_' type '_' namendo,'_vs_', namexo '_threshold' ],'lpmat','lpmat0','istable','y0','x0','xx0','iy','iyc')
lpmat=[]; lpmat0=[]; istable=[];
if length(iy)<=10 || length(iyc)<=10
icheck = []; % do the generic plot in any case
end
else
icheck=[];
end
if isempty(icheck)
if length(iy)<=10
if isempty(iy)
disp(['There are NO MC samples in the desired range [' num2str(threshold) ']!'])
else
disp(['There are TOO FEW (<=10) MC samples in the desired range [' num2str(threshold) ']!'])
end
elseif length(iyc)<=10
if isempty(iyc)
disp(['ALL MC samples are in the desired range [' num2str(threshold) ']!'])
else
disp(['Almost ALL MC samples are in the desired range [' num2str(threshold) ']!'])
disp('There are TOO FEW (<=10) MC samples OUTSIDE the desired range!')
end
end
atitle0=['Monte Carlo Filtering for ',namendo,' vs ', namexo];
options_mcf.title = atitle0;
indmcf = redform_mcf(y0, x0, options_mcf, options_);
@ -280,6 +297,8 @@ for j = 1:length(anamendo)
end
end
else
disp(['This entry in the shock matrix is CONSTANT = ' num2str(mean(y0),3)])
end
end
end
@ -369,11 +388,28 @@ for j = 1:length(anamendo)
istable=[1:length(iy)];
save([xdir,filesep, fname_ '_' type '_' namendo,'_vs_', namlagendo '_threshold' ],'lpmat','lpmat0','istable','y0','x0','xx0','iy','iyc')
lpmat=[]; lpmat0=[]; istable=[];
if length(iy)<=10 || length(iyc)<=10,
icheck = []; % do the generic plot in any case
end
else
icheck = [];
end
if isempty(icheck)
if length(iy)<=10
if isempty(iy)
disp(['There are NO MC samples in the desired range [' num2str(threshold) ']!'])
else
disp(['There are TOO FEW (<=10) MC samples in the desired range [' num2str(threshold) ']!'])
end
elseif length(iyc)<=10
if isempty(iyc)
disp(['ALL MC samples are in the desired range [' num2str(threshold) ']!'])
else
disp(['Almost ALL MC samples are in the desired range [' num2str(threshold) ']!'])
disp('There are TOO FEW (<=10) MC samples OUTSIDE the desired range!')
end
end
atitle0=['Monte Carlo Filtering for ',namendo,' vs ', namlagendo];
options_mcf.title = atitle0;
indmcf = redform_mcf(y0, x0, options_mcf, options_);
@ -417,6 +453,8 @@ for j = 1:length(anamendo)
end
end
else
disp(['This entry in the transition matrix is CONSTANT = ' num2str(mean(y0),3)])
end
end
end
@ -736,6 +774,11 @@ for jt=1:10
end
[proba, dproba] = stab_map_1(x0, indy{1}, indy{end}, [],0);
indmcf=find(proba<options_mcf.pvalue_ks);
if isempty(indmcf)
[tmp,jtmp] = sort(proba,2,'ascend');
indmcf = jtmp(1);
% indmcf = jtmp(1:min(2,length(proba)));
end
[tmp,jtmp] = sort(proba(indmcf),2,'ascend');
indmcf = indmcf(jtmp);
nbr_par = length(indmcf);

6
matlab/identification_analysis.m
View File

@ -28,7 +28,7 @@ function [ide_hess, ide_moments, ide_model, ide_lre, derivatives_info, info, opt
% SPECIAL REQUIREMENTS
% None
% Copyright (C) 2008-2017 Dynare Team
% Copyright (C) 2008-2018 Dynare Team
%
% This file is part of Dynare.
%
@ -152,10 +152,10 @@ if info(1)==0
end
analytic_derivation = options_.analytic_derivation;
options_.analytic_derivation = -2;
info = stoch_simul(char(options_.varobs));
info = stoch_simul(options_.varobs);
dataset_ = dseries(oo_.endo_simul(options_.varobs_id,100+1:end)',dates('1Q1'), options_.varobs);
derivatives_info.no_DLIK=1;
%bounds = prior_bounds(bayestopt_, options_.prior_trunc);
bounds = prior_bounds(bayestopt_, options_.prior_trunc); %reset bounds as lb and ub must only be operational during mode-finding
[fval,info,cost_flag,DLIK,AHess,ys,trend_coeff,M_,options_,bayestopt_,oo_] = dsge_likelihood(params',dataset_,dataset_info,options_,M_,estim_params_,bayestopt_,bounds,oo_,derivatives_info);
% fval = DsgeLikelihood(xparam1,data_info,options_,M_,estim_params_,bayestopt_,oo_);
options_.analytic_derivation = analytic_derivation;

9
matlab/initial_condition_decomposition.m
View File

@ -128,6 +128,11 @@ M_.exo_nbr = M_.endo_nbr;
options_.plot_shock_decomp.realtime=0;
options_.plot_shock_decomp.screen_shocks=1;
options_.plot_shock_decomp.use_shock_groups = '';
options_.plot_shock_decomp.fig_names='initval';
plot_shock_decomposition(M_, oo, options_, varlist);
fig_name = options_.plot_shock_decomp.fig_name;
if ~isempty(fig_name)
options_.plot_shock_decomp.fig_name=[fig_name '_initval'];
else
options_.plot_shock_decomp.fig_name='initval';
end
plot_shock_decomposition(M_,oo,options_,varlist);
% end

4
matlab/kalman/likelihood/kalman_filter_d.m
View File

@ -36,7 +36,7 @@ function [dLIK,dlik,a,Pstar] = kalman_filter_d(Y, start, last, a, Pinf, Pstar, k
% Durbin/Koopman (2012): "Time Series Analysis by State Space Methods", Oxford University Press,
% Second Edition, Ch. 5 and 7.2
% Copyright (C) 2004-2017 Dynare Team
% Copyright (C) 2004-2018 Dynare Team
%
% This file is part of Dynare.
%
@ -65,7 +65,7 @@ dLIK = Inf; % Default value of the log likelihood.
oldK = Inf;
s = 0;
while rank(Pinf,diffuse_kalman_tol) && (t<=last)
while rank(Z*Pinf*Z',diffuse_kalman_tol) && (t<=last)
s = t-start+1;
v = Y(:,t)-Z*a; %get prediction error v^(0) in (5.13) DK (2012)
Finf = Z*Pinf*Z'; % (5.7) in DK (2012)

7
matlab/kalman/likelihood/univariate_kalman_filter_d.m
View File

@ -87,7 +87,7 @@ function [dLIK, dlikk, a, Pstar, llik] = univariate_kalman_filter_d(data_index,
% Series Analysis by State Space Methods", Oxford University Press,
% Second Edition, Ch. 5, 6.4 + 7.2.5
% Copyright (C) 2004-2017 Dynare Team
% Copyright (C) 2004-2018 Dynare Team
%
% This file is part of Dynare.
%
@ -154,7 +154,7 @@ while newRank && (t<=last)
end
end
if newRank
oldRank = rank(Pinf,diffuse_kalman_tol);
oldRank = rank(Z*Pinf*Z',diffuse_kalman_tol);
else
oldRank = 0;
end
@ -162,10 +162,11 @@ while newRank && (t<=last)
Pstar = T*Pstar*T'+QQ;
Pinf = T*Pinf*T';
if newRank
newRank = rank(Pinf,diffuse_kalman_tol);
newRank = rank(Z*Pinf*Z',diffuse_kalman_tol);
end
if oldRank ~= newRank
disp('univariate_diffuse_kalman_filter:: T does influence the rank of Pinf!')
disp('This may happen for models with order of integration >1.')
end
t = t+1;
end

83
matlab/missing/nanmean/nanmean.m
View File

@ -1,36 +1,20 @@
function y = nanmean(x)
% Computes the mean of each column of a matrix with some NaNs.
function y = nanmean(x, dim)
%@info:
%! @deftypefn {Function File} {@var{y} =} nanmean (@var{x})
%! @anchor{nanmean}
%! @sp 1
%! Computes the mean of each column of a matrix with some NaNs.
%! @sp 2
%! @strong{Inputs}
%! @table @ @var
%! @item x
%! Matlab matrix (T-by-N).
%! @end table
%! @sp 2
%! @strong{Outputs}
%! @table @ @var
%! @item y
%! Matlab vector (1-by-N), the mean.
%! @end table
%! @sp 2
%! @strong{This function is called by:}
%! @sp 1
%! @ref{compute_cova}, @ref{compute_acov}, @ref{compute_std}, @ref{nandemean}
%! @sp 2
%! @strong{This function calls:}
%! @sp 1
%! @ref{ndim}
%!
%! @end deftypefn
%@eod:
% Returns the mean of a matrix with some NaNs.
%
% INPUTS
% - x [double] m*n matrix
% - dim [integer] scalar, dimension along which the mean has to be computed.
%
% OUTPUTS
% - y [double] 1*n vector (if dim=1) or m*1 vector (if dim=2).
%
% REMARKS
% (1) Default value for dim is the first non singleton dimension.
% (2) Works with vector and matrices, not implemented for arrays with more
% than two dimensions.
% Copyright (C) 2011 Dynare Team
% Copyright (C) 2011-2018 Dynare Team
%
% This file is part of Dynare.
%
@ -47,16 +31,39 @@ function y = nanmean(x)
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
% AUTHOR(S) stephane DOT adjemian AT univ DASH lemans DOT fr
if nargin<2
% By default dim is the first non singleton dimension
nonsingletondims = find(find(size(x)>1));
if ~isempty(nonsingletondims)
dim = nonsingletondims(1);
else
dim = NaN;
end
end
switch ndim(x)
switch ndims(x)
case 1
y = mean(x(find(~isnan(x))));
if isnan(dim)
y = x;
else
y = mean(x(find(~isnan(x))), dim);
end
case 2
y = NaN(1,size(x,2));
for i = 1:size(x,2)
y(i) = mean(x(find(~isnan(x(:,i))),i));
if isnan(dim)
y = x;
else
if isequal(dim, 1)
y = NaN(1, size(x, 2));
for i = 1:size(x, 2)
y(i) = mean(x(find(~isnan(x(:,i))),i));
end
else
y = NaN(size(x, 1), 1);
for i = 1:size(x, 1)
y(i) = mean(x(i, find(~isnan(x(i,:)))));
end
end
end
otherwise
error('descriptive_statistics::nanmean:: This function is not implemented for arrays with dimension greater than two!')
error('This function is not implemented for arrays with dimension greater than two!')
end

14
matlab/missing_DiffuseKalmanSmootherH1_Z.m
View File

@ -49,7 +49,7 @@ function [alphahat,epsilonhat,etahat,atilde,P,aK,PK,decomp,V] = missing_DiffuseK
% Durbin/Koopman (2012): "Time Series Analysis by State Space Methods", Oxford University Press,
% Second Edition, Ch. 5
% Copyright (C) 2004-2017 Dynare Team
% Copyright (C) 2004-2018 Dynare Team
%
% This file is part of Dynare.
%
@ -134,9 +134,9 @@ while rank(Pinf(:,:,t+1),diffuse_kalman_tol) && t<smpl
return
else %rank of F_{\infty,t} is 0
Finf_singular(1,t) = 1;
Fstar(:,:,t) = ZZ*Pstar(:,:,t)*ZZ' + H(di,di); % (5.7) in DK (2012)
if rcond(Fstar(:,:,t)) < kalman_tol %F_{*} is singular
if ~all(abs(Fstar(:,:,t))<kalman_tol)
Fstar(di,di,t) = ZZ*Pstar(:,:,t)*ZZ' + H(di,di); % (5.7) in DK (2012)
if rcond(Fstar(di,di,t)) < kalman_tol %F_{*} is singular
if ~all(all(abs(Fstar(di,di,t))<kalman_tol))
% The univariate diffuse kalman filter should be used.
alphahat = Inf;
return
@ -146,12 +146,12 @@ while rank(Pinf(:,:,t+1),diffuse_kalman_tol) && t<smpl
Pinf(:,:,t+1) = T*Pinf(:,:,t)*transpose(T);
end
else
iFstar(:,:,t) = inv(Fstar(:,:,t));
Kstar(:,:,t) = Pstar(:,:,t)*ZZ'*iFstar(:,:,t); %(5.15) of DK (2012) with Kstar=T^{-1}*K^(0)
iFstar(di,di,t) = inv(Fstar(di,di,t));
Kstar(:,di,t) = Pstar(:,:,t)*ZZ'*iFstar(di,di,t); %(5.15) of DK (2012) with Kstar=T^{-1}*K^(0)
Pinf(:,:,t+1) = T*Pinf(:,:,t)*transpose(T); % DK (2012), 5.16
Lstar(:,:,t) = T - T*Kstar(:,di,t)*ZZ; %L^(0) in DK (2012), eq. 5.12
Pstar(:,:,t+1) = T*Pstar(:,:,t)*Lstar(:,:,t)'+QQ; % (5.17) DK (2012)
a(:,t+1) = T*(a(:,t)+Kstar(:,:,t)*v(:,t)); % (5.13) DK (2012)
a(:,t+1) = T*(a(:,t)+Kstar(:,di,t)*v(di,t)); % (5.13) DK (2012)
end
end
else

3
matlab/missing_DiffuseKalmanSmootherH3_Z.m
View File

@ -56,7 +56,7 @@ function [alphahat,epsilonhat,etahat,a,P,aK,PK,decomp,V] = missing_DiffuseKalman
% Models", S.J. Koopman and J. Durbin (2003), in Journal of Time Series
% Analysis, vol. 24(1), pp. 85-98.
% Copyright (C) 2004-2017 Dynare Team
% Copyright (C) 2004-2018 Dynare Team
%
% This file is part of Dynare.
%
@ -184,6 +184,7 @@ while newRank && t < smpl
end
if oldRank ~= newRank
disp('univariate_diffuse_kalman_filter:: T does influence the rank of Pinf!')
disp('This may happen for models with order of integration >1.')
end
end

14
matlab/mode_check.m
View File

@ -111,14 +111,22 @@ for plt = 1:nbplt
end
end
xx = x;
if x(kk)~=0
if x(kk)~=0 || ~isinf(BoundsInfo.lb(kk)) || ~isinf(BoundsInfo.lb(kk))
l1 = max(BoundsInfo.lb(kk),(1-sign(x(kk))*ll)*x(kk)); m1 = 0; %lower bound
l2 = min(BoundsInfo.ub(kk),(1+sign(x(kk))*ll)*x(kk)); %upper bound
else
%size info for 0 parameter is missing, use prior standard
%deviation
l1 = max(BoundsInfo.lb(kk),-BayesInfo.p2(kk)); m1 = 0; %lower bound
l2 = min(BoundsInfo.ub(kk),BayesInfo.p2(kk)); %upper bound
upper_bound=BoundsInfo.lb(kk);
if isinf(upper_bound)
upper_bound=-1e-6*DynareOptions.huge_number;
end
lower_bound=BoundsInfo.ub(kk);
if isinf(lower_bound)
lower_bound=-1e-6*DynareOptions.huge_number;
end
l1 = max(lower_bound,-BayesInfo.p2(kk)); m1 = 0; %lower bound
l2 = min(upper_bound,BayesInfo.p2(kk)); %upper bound
end
binding_lower_bound=0;
binding_upper_bound=0;

2
matlab/model_diagnostics.m
View File

@ -154,7 +154,7 @@ for b=1:nb
break
end
end
disp(endo_names{k})
fprintf('%s\n',endo_names{k})
end
neq = null(jacob');
n_rel = size(neq,2);

2
matlab/modules/dseries

@ -1 +1 @@
Subproject commit cad90545bd355f671c112fd072b9be1545cbb2c4
Subproject commit 656e63e360d9fa979f932b601c46b1c1ea8aa1a4

117
matlab/parallel/AnalyseComputationalEnvironment.m
View File

@ -251,10 +251,7 @@ for Node=1:length(DataInput) % To obtain a recoursive function remove the 'for'
% computer!
if Environment
% This check can be removed ... according to the dynare parser
% strategy.
% This check can be removed ... according to the dynare parser strategy.
if isempty(DataInput(Node).RemoteDirectory)
disp('The field RemoteDirectory is empty!')
skipline()
@ -263,10 +260,7 @@ for Node=1:length(DataInput) % To obtain a recoursive function remove the 'for'
ErrorCode=5;
return
end
% This check can be removed ... according to the dynare parser
% strategy.
% This check can be removed ... according to the dynare parser strategy.
if (~isempty(DataInput(Node).RemoteDrive))
disp('[WARNING] The fields RemoteDrive should be empty under unix or mac!')
skipline()
@ -276,35 +270,14 @@ for Node=1:length(DataInput) % To obtain a recoursive function remove the 'for'
skipline(2)
ErrorCode=5;
end
si2=[];
de2=[];
if ~isempty(DataInput(Node).Port)
ssh_token = ['-p ',DataInput(Node).Port];
else
ssh_token = '';
end
[si2 de2]=system(['ssh ',ssh_token,' ',DataInput(Node).UserName,'@',DataInput(Node).ComputerName,' ls ',DataInput(Node).RemoteDirectory,'/',RemoteTmpFolder,'/']);
if (si2)
disp ('Remote Directory does not exist or is not reachable!')
skipline()
disp('ErrorCode 5.')
skipline(2)
ErrorCode=5;
return
end
disp('Check on RemoteDirectory Variable ..... Ok!')
skipline(2)
disp('Check on RemoteDrive Variable ..... Ok!')
skipline(2)
command_string = ['ssh ',ssh_token,' ',DataInput(Node).UserName,'@',DataInput(Node).ComputerName,' ls ',DataInput(Node).RemoteDirectory,'/',RemoteTmpFolder,'/'];
else
% This check can be removed ... according to the dynare parser
% strategy.
% This check can be removed ... according to the dynare parser strategy.
if (isempty(DataInput(Node).RemoteDrive)||isempty(DataInput(Node).RemoteDirectory))
disp('Remote RemoteDrive and/or RemoteDirectory is/are empty!')
skipline()
@ -313,32 +286,31 @@ for Node=1:length(DataInput) % To obtain a recoursive function remove the 'for'
ErrorCode=5;
return
end
si2=[];
de2=[];
[si2 de2]=system(['dir \\',DataInput(Node).ComputerName,'\',DataInput(Node).RemoteDrive,'$\',DataInput(Node).RemoteDirectory,'\',RemoteTmpFolder]);
if (si2)
disp ('Remote Directory does not exist or it is not reachable!')
skipline()
disp('ErrorCode 5.')
skipline(2)
ErrorCode=5;
return
end
disp('Check on RemoteDirectory Variable ..... Ok!')
skipline(2)
disp('Check on RemoteDrive Variable ..... Ok!')
skipline(2)
command_string = ['dir \\',DataInput(Node).ComputerName,'\',DataInput(Node).RemoteDrive,'$\',DataInput(Node).RemoteDirectory,'\',RemoteTmpFolder];
end
[si2, de2] = system(command_string);
% Now we verify if it possible to exchange data with the remote
% computer:
if (si2)
disp ('Remote Directory does not exist or is not reachable!')
skipline()
disp('ErrorCode 5.')
skipline(2)
disp('The command causing the error was:')
disp(command_string)
disp('The system returned:')
disp(de2)
skipline(2)
ErrorCode=5;
return
end
disp('Check on RemoteDirectory Variable ..... Ok!')
skipline(2)
disp('Check on RemoteDrive Variable ..... Ok!')
skipline(2)
% Now we verify if it possible to exchange data with the remote computer.
% Build a command file to test the matlab execution and dynare path ...
@ -506,13 +478,8 @@ for Node=1:length(DataInput) % To obtain a recoursive function remove the 'for'
% Now we verify if it is possible delete remote computational traces!
dynareParallelRmDir(RemoteTmpFolder,DataInput(Node));
si3=[];
si3=dynareParallelDir('Tracing.m', RemoteTmpFolder,DataInput(Node));
si3 = dynareParallelDir('Tracing.m', RemoteTmpFolder,DataInput(Node));
if (isempty(si3))
disp ('Check on Delete Remote Computational Traces ..... Ok!')
skipline(2)
@ -550,22 +517,18 @@ for Node=1:length(DataInput) % To obtain a recoursive function remove the 'for'
% We look for the information on local computer hardware.
si0=[];
de0=[];
Environment1=Environment;
Environment1 = Environment;
disp('Checking Hardware please wait ...');
if (DataInput(Node).Local == 1)
if Environment
if ~ismac
[si0 de0]=system('grep processor /proc/cpuinfo');
[si0, de0] = system('grep processor /proc/cpuinfo');
else
[si0 de0]=system('sysctl -n hw.ncpu');
Environment1=2;
[si0, de0] = system('sysctl -n hw.ncpu');
Environment1 = 2;
end
else
[si0 de0]=system(['psinfo \\']);
[si0, de0] = system(['psinfo \\']);
end
else
if Environment
@ -576,19 +539,27 @@ for Node=1:length(DataInput) % To obtain a recoursive function remove the 'for'
end
if OStargetUnix
if RemoteEnvironment ==1
[si0 de0]=system(['ssh ',ssh_token,' ',DataInput(Node).UserName,'@',DataInput(Node).ComputerName,' grep processor /proc/cpuinfo']);
command_string = ['ssh ',ssh_token,' ',DataInput(Node).UserName,'@',DataInput(Node).ComputerName,' grep processor /proc/cpuinfo'];
else % it is MAC
[si0 de0]=system(['ssh ',ssh_token,' ',DataInput(Node).UserName,'@',DataInput(Node).ComputerName,' sysctl -n hw.ncpu']);
Environment1=2;
command_string = ['ssh ',ssh_token,' ',DataInput(Node).UserName,'@',DataInput(Node).ComputerName,' sysctl -n hw.ncpu'];
Environment1 = 2;
end
else
[si0 de0]=system(['ssh ',ssh_token,' ',DataInput(Node).UserName,'@',DataInput(Node).ComputerName,' psinfo']);
command_string = ['ssh ',ssh_token,' ',DataInput(Node).UserName,'@',DataInput(Node).ComputerName,' psinfo'];
end
else
[si0 de0]=system(['psinfo \\',DataInput(Node).ComputerName,' -u ',DataInput(Node).UserName,' -p ',DataInput(Node).Password]);
command_string = ['psinfo \\',DataInput(Node).ComputerName,' -u ',DataInput(Node).UserName,' -p ',DataInput(Node).Password];
end
[si0, de0] = system(command_string);
end
if (si0)
disp('The command causing the error was:')
disp(command_string)
disp('The system returned:')
disp(de0)
skipline(2)
end
RealCPUnbr='';
% keyboard;

24
matlab/parallel/masterParallel.m
View File

@ -347,9 +347,9 @@ for j=1:totCPU
end
else % Hybrid computing Matlab(Master)->Octave(Slaves) and Vice Versa!
if regexpi([Parallel(indPC).MatlabOctavePath], 'octave')
command1=['psexec -d -W "',DyMo, '" -a ',my_affinity,' -low ',Parallel(indPC).MatlabOctavePath,' -f --eval "default_save_options(''-v7''); addpath(''',Parallel(indPC).DynarePath,'''), dynareroot = dynare_config(); fParallel(',int2str(offset+1),',',int2str(sum(nBlockPerCPU(1:j))),',',int2str(j),',',int2str(indPC),',''',fname,''')"'];
command1=['psexec -accepteula -d -W "',DyMo, '" -a ',my_affinity,' -low ',Parallel(indPC).MatlabOctavePath,' -f --eval "default_save_options(''-v7''); addpath(''',Parallel(indPC).DynarePath,'''), dynareroot = dynare_config(); fParallel(',int2str(offset+1),',',int2str(sum(nBlockPerCPU(1:j))),',',int2str(j),',',int2str(indPC),',''',fname,''')"'];
else
command1=['psexec -d -W "',DyMo, '" -a ',my_affinity,' -low ',Parallel(indPC).MatlabOctavePath,' -nosplash -nodesktop -minimize ',compThread,' -r "addpath(''',Parallel(indPC).DynarePath,'''), dynareroot = dynare_config(); fParallel(',int2str(offset+1),',',int2str(sum(nBlockPerCPU(1:j))),',',int2str(j),',',int2str(indPC),',''',fname,''')"'];
command1=['psexec -accepteula -d -W "',DyMo, '" -a ',my_affinity,' -low ',Parallel(indPC).MatlabOctavePath,' -nosplash -nodesktop -minimize ',compThread,' -r "addpath(''',Parallel(indPC).DynarePath,'''), dynareroot = dynare_config(); fParallel(',int2str(offset+1),',',int2str(sum(nBlockPerCPU(1:j))),',',int2str(j),',',int2str(indPC),',''',fname,''')"'];
end
end
else % 0.2 Parallel(indPC).Local==0: Run using network on remote machine or also on local machine.
@ -387,20 +387,20 @@ for j=1:totCPU
if ~strcmpi(Parallel(indPC).ComputerName,MasterName) % 0.3 Run on a remote machine!
% Hybrid computing Matlab(Master)-> Octave(Slaves) and Vice Versa!
if regexpi([Parallel(indPC).MatlabOctavePath], 'octave')
command1=['psexec \\',Parallel(indPC).ComputerName,' -d -e -u ',Parallel(indPC).UserName,' -p ',Parallel(indPC).Password,' -W "',Parallel(indPC).RemoteDrive,':\',Parallel(indPC).RemoteDirectory,'\',PRCDir,'\" -a ',my_affinity, ...
command1=['psexec \\',Parallel(indPC).ComputerName,' -accepteula -d -e -u ',Parallel(indPC).UserName,' -p ',Parallel(indPC).Password,' -W "',Parallel(indPC).RemoteDrive,':\',Parallel(indPC).RemoteDirectory,'\',PRCDir,'\" -a ',my_affinity, ...
' -low ',Parallel(indPC).MatlabOctavePath,' -f --eval "default_save_options(''-v7''); addpath(''',Parallel(indPC).DynarePath,'''), dynareroot = dynare_config(); fParallel(',int2str(offset+1),',',int2str(sum(nBlockPerCPU(1:j))),',',int2str(j),',',int2str(indPC),',''',fname,''')"'];
else
command1=['psexec \\',Parallel(indPC).ComputerName,' -d -e -u ',Parallel(indPC).UserName,' -p ',Parallel(indPC).Password,' -W "',Parallel(indPC).RemoteDrive,':\',Parallel(indPC).RemoteDirectory,'\',PRCDir,'\" -a ',my_affinity, ...
command1=['psexec \\',Parallel(indPC).ComputerName,' -accepteula -d -e -u ',Parallel(indPC).UserName,' -p ',Parallel(indPC).Password,' -W "',Parallel(indPC).RemoteDrive,':\',Parallel(indPC).RemoteDirectory,'\',PRCDir,'\" -a ',my_affinity, ...
' -low ',Parallel(indPC).MatlabOctavePath,' -nosplash -nodesktop -minimize ',compThread,' -r "addpath(''',Parallel(indPC).DynarePath,'''), dynareroot = dynare_config(); fParallel(',int2str(offset+1),',',int2str(sum(nBlockPerCPU(1:j))),',',int2str(j),',',int2str(indPC),',''',fname,''')"'];
end
else % 0.4 Run on the local machine via the network
% Hybrid computing Matlab(Master)->Octave(Slaves) and Vice Versa!
if regexpi([Parallel(indPC).MatlabOctavePath], 'octave')
command1=['psexec \\',Parallel(indPC).ComputerName,' -d -e -W "',Parallel(indPC).RemoteDrive,':\',Parallel(indPC).RemoteDirectory,'\',PRCDir,'\" -a ',my_affinity, ...
command1=['psexec \\',Parallel(indPC).ComputerName,' -accepteula -d -e -W "',Parallel(indPC).RemoteDrive,':\',Parallel(indPC).RemoteDirectory,'\',PRCDir,'\" -a ',my_affinity, ...
' -low ',Parallel(indPC).MatlabOctavePath,' -f --eval "default_save_options(''-v7''); addpath(''',Parallel(indPC).DynarePath,'''), dynareroot = dynare_config(); fParallel(',int2str(offset+1),',',int2str(sum(nBlockPerCPU(1:j))),',',int2str(j),',',int2str(indPC),',''',fname,''')"'];
else
command1=['psexec \\',Parallel(indPC).ComputerName,' -d -e -W "',Parallel(indPC).RemoteDrive,':\',Parallel(indPC).RemoteDirectory,'\',PRCDir,'\" -a ',my_affinity, ...
command1=['psexec \\',Parallel(indPC).ComputerName,' -accepteula -d -e -W "',Parallel(indPC).RemoteDrive,':\',Parallel(indPC).RemoteDirectory,'\',PRCDir,'\" -a ',my_affinity, ...
' -low ',Parallel(indPC).MatlabOctavePath,' -nosplash -nodesktop -minimize ',compThread,' -r "addpath(''',Parallel(indPC).DynarePath,'''), dynareroot = dynare_config(); fParallel(',int2str(offset+1),',',int2str(sum(nBlockPerCPU(1:j))),',',int2str(j),',',int2str(indPC),',''',fname,''')"'];
end
end
@ -418,9 +418,9 @@ for j=1:totCPU
end
else % Hybrid computing Matlab(Master)->Octave(Slaves) and Vice Versa!
if regexpi([Parallel(indPC).MatlabOctavePath], 'octave')
command1=['psexec -d -W "',DyMo, '" -a ',my_affinity,' -low ',Parallel(indPC).MatlabOctavePath,' -f --eval "default_save_options(''-v7'');addpath(''',Parallel(indPC).DynarePath,'''), dynareroot = dynare_config(); slaveParallel(',int2str(j),',',int2str(indPC),')"'];
command1=['psexec -accepteula -d -W "',DyMo, '" -a ',my_affinity,' -low ',Parallel(indPC).MatlabOctavePath,' -f --eval "default_save_options(''-v7'');addpath(''',Parallel(indPC).DynarePath,'''), dynareroot = dynare_config(); slaveParallel(',int2str(j),',',int2str(indPC),')"'];
else
command1=['psexec -d -W "',DyMo, '" -a ',my_affinity,' -low ',Parallel(indPC).MatlabOctavePath,' -nosplash -nodesktop -minimize ',compThread,' -r "addpath(''',Parallel(indPC).DynarePath,'''), dynareroot = dynare_config(); slaveParallel(',int2str(j),',',int2str(indPC),')"'];
command1=['psexec -accepteula -d -W "',DyMo, '" -a ',my_affinity,' -low ',Parallel(indPC).MatlabOctavePath,' -nosplash -nodesktop -minimize ',compThread,' -r "addpath(''',Parallel(indPC).DynarePath,'''), dynareroot = dynare_config(); slaveParallel(',int2str(j),',',int2str(indPC),')"'];
end
end
elseif Parallel(indPC).Local==0 % 1.2 Run using network on remote machine or also on local machine.
@ -462,19 +462,19 @@ for j=1:totCPU
if ~strcmpi(Parallel(indPC).ComputerName,MasterName) % 1.3 Run on a remote machine.
% Hybrid computing Matlab(Master)->Octave(Slaves) and Vice Versa!
if regexpi([Parallel(indPC).MatlabOctavePath], 'octave')
command1=['psexec \\',Parallel(indPC).ComputerName,' -d -e -u ',Parallel(indPC).UserName,' -p ',Parallel(indPC).Password,' -W "',Parallel(indPC).RemoteDrive,':\',Parallel(indPC).RemoteDirectory,'\',PRCDir,'\" -a ',my_affinity, ...
command1=['psexec \\',Parallel(indPC).ComputerName,' -accepteula -d -e -u ',Parallel(indPC).UserName,' -p ',Parallel(indPC).Password,' -W "',Parallel(indPC).RemoteDrive,':\',Parallel(indPC).RemoteDirectory,'\',PRCDir,'\" -a ',my_affinity, ...
' -low ',Parallel(indPC).MatlabOctavePath,' -f --eval "default_save_options(''-v7'');addpath(''',Parallel(indPC).DynarePath,'''), dynareroot = dynare_config(); slaveParallel(',int2str(j),',',int2str(indPC),')"'];
else
command1=['psexec \\',Parallel(indPC).ComputerName,' -d -e -u ',Parallel(indPC).UserName,' -p ',Parallel(indPC).Password,' -W "',Parallel(indPC).RemoteDrive,':\',Parallel(indPC).RemoteDirectory,'\',PRCDir,'\" -a ',my_affinity, ...
command1=['psexec \\',Parallel(indPC).ComputerName,' -accepteula -d -e -u ',Parallel(indPC).UserName,' -p ',Parallel(indPC).Password,' -W "',Parallel(indPC).RemoteDrive,':\',Parallel(indPC).RemoteDirectory,'\',PRCDir,'\" -a ',my_affinity, ...
' -low ',Parallel(indPC).MatlabOctavePath,' -nosplash -nodesktop -minimize ',compThread,' -r "addpath(''',Parallel(indPC).DynarePath,'''), dynareroot = dynare_config(); slaveParallel(',int2str(j),',',int2str(indPC),')"'];
end
else % 1.4 Run on the local machine via the network.
% Hybrid computing Matlab(Master)->Octave(Slaves) and Vice Versa!
if regexpi([Parallel(indPC).MatlabOctavePath], 'octave')
command1=['psexec \\',Parallel(indPC).ComputerName,' -d -e -W "',Parallel(indPC).RemoteDrive,':\',Parallel(indPC).RemoteDirectory,'\',PRCDir,'\" -a ',my_affinity, ...
command1=['psexec \\',Parallel(indPC).ComputerName,' -accepteula -d -e -W "',Parallel(indPC).RemoteDrive,':\',Parallel(indPC).RemoteDirectory,'\',PRCDir,'\" -a ',my_affinity, ...
' -low ',Parallel(indPC).MatlabOctavePath,' -f --eval "default_save_options(''-v7''); addpath(''',Parallel(indPC).DynarePath,'''), dynareroot = dynare_config(); slaveParallel(',int2str(j),',',int2str(indPC),')"'];
else
command1=['psexec \\',Parallel(indPC).ComputerName,' -d -e -W "',Parallel(indPC).RemoteDrive,':\',Parallel(indPC).RemoteDirectory,'\',PRCDir,'\" -a ',my_affinity, ...
command1=['psexec \\',Parallel(indPC).ComputerName,' -accepteula -d -e -W "',Parallel(indPC).RemoteDrive,':\',Parallel(indPC).RemoteDirectory,'\',PRCDir,'\" -a ',my_affinity, ...
' -low ',Parallel(indPC).MatlabOctavePath,' -nosplash -nodesktop -minimize ',compThread,' -r "addpath(''',Parallel(indPC).DynarePath,'''), dynareroot = dynare_config(); slaveParallel(',int2str(j),',',int2str(indPC),')"'];
end
end

2
matlab/perfect-foresight-models/sim1_linear.m
View File

@ -116,7 +116,7 @@ x = repmat(transpose(steadystate_x), 1+M.maximum_exo_lag+M.maximum_exo_lead, 1);
[d1, jacobian] = dynamicmodel(z, x, params, steadystate_y, M.maximum_exo_lag+1);
% Check that the dynamic model was evaluated at the steady state.
if max(abs(d1))>1e-12
if max(abs(d1))>options.solve_tolf
error('Jacobian is not evaluated at the steady state!')
end

2
matlab/posterior_sampler.m
View File

@ -127,7 +127,7 @@ else
% which files have to be copied to run remotely
NamFileInput(1,:) = {'',[ModelName '_static.m']};
NamFileInput(2,:) = {'',[ModelName '_dynamic.m']};
if M.set_auxiliary_variables
if M_.set_auxiliary_variables
NamFileInput(3,:) = {'',[M_.fname '_set_auxiliary_variables.m']};
end
if options_.steadystate_flag

4
matlab/read_variables.m
View File

@ -91,7 +91,7 @@ switch (extension)
case { '.xls', '.xlsx' }
[freq,init,data,varlist] = load_xls_file_data(fullname,xls_sheet,xls_range);
for dyn_i_01=1:var_size_01
iv = strmatch(strtrim(var_names_01(dyn_i_01,:)),varlist,'exact');
iv = strmatch(strtrim(var_names_01{dyn_i_01}),varlist,'exact');
if ~isempty(iv)
dyn_tmp_01 = [data(:,iv)]';
if length(dyn_tmp_01) > dyn_size_01 && dyn_size_01 > 0
@ -101,7 +101,7 @@ switch (extension)
dyn_data_01(:,dyn_i_01) = dyn_tmp_01;
else
cd(old_pwd)
error([strtrim(var_names_01(dyn_i_01,:)) ' not found in ' fullname])
error([strtrim(var_names_01{dyn_i_01}) ' not found in ' fullname])
end
end
case '.csv'

58
matlab/smoother2histval.m
View File

@ -193,36 +193,38 @@ end
% Handle auxiliary variables for lags (both on endogenous and exogenous)
for i = 1:length(M_.aux_vars)
if M_.aux_vars(i).type ~= 1 && M_.aux_vars(i).type ~= 3
continue
end
if M_.aux_vars(i).type == 1
% Endogenous
orig_var = M_.endo_names{M_.aux_vars(i).orig_index};
else
% Exogenous
orig_var = M_.exo_names{M_.aux_vars(i).orig_index};
end
[m, k] = ismember(orig_var, outvars);
if m
if ~isempty(strmatch(invars{k}, M_.endo_names))
s = getfield(smoothedvars, invars{k});
else
s = getfield(smoothedshocks, invars{k});
if ~ ismember(M_.endo_names{M_.aux_vars(i).endo_index},invars)
if M_.aux_vars(i).type ~= 1 && M_.aux_vars(i).type ~= 3
continue
end
l = M_.aux_vars(i).orig_lead_lag;
if period-M_.maximum_endo_lag+1+l < 1
error('The period that you indicated is too small to construct initial conditions')
end
j = M_.aux_vars(i).endo_index;
v = s((period-M_.maximum_endo_lag+1+l):(period+l)); %+steady_state(j);
if ~isfield(opts, 'outfile')
M_.endo_histval(j, :) = v;
if M_.aux_vars(i).type == 1
% Endogenous
orig_var = M_.endo_names{M_.aux_vars(i).orig_index};
else
% When saving to a file, x(-2) is in the variable called "x_l2"
lead_lag = num2str(l);
lead_lag = regexprep(lead_lag, '-', 'l');
o = setfield(o, [ orig_var '_' lead_lag ], v);
% Exogenous
orig_var = M_.exo_names{M_.aux_vars(i).orig_index};
end
[m, k] = ismember(orig_var, outvars);
if m
if ~isempty(strmatch(invars{k}, M_.endo_names))
s = getfield(smoothedvars, invars{k});
else
s = getfield(smoothedshocks, invars{k});
end
l = M_.aux_vars(i).orig_lead_lag;
if period-M_.maximum_endo_lag+1+l < 1
error('The period that you indicated is too small to construct initial conditions')
end
j = M_.aux_vars(i).endo_index;
v = s((period-M_.maximum_endo_lag+1+l):(period+l)); %+steady_state(j);
if ~isfield(opts, 'outfile')
M_.endo_histval(j, :) = v;
else
% When saving to a file, x(-2) is in the variable called "x_l2"
lead_lag = num2str(l);
lead_lag = regexprep(lead_lag, '-', 'l');
o = setfield(o, [ orig_var '_' lead_lag ], v);
end
end
end
end

5
matlab/stochastic_solvers.m
View File

@ -29,7 +29,7 @@ function [dr,info] = stochastic_solvers(dr,task,M_,options_,oo_)
% none.
%
% Copyright (C) 1996-2017 Dynare Team
% Copyright (C) 1996-2018 Dynare Team
%
% This file is part of Dynare.
%
@ -57,6 +57,9 @@ if M_.hessian_eq_zero && local_order~=1
local_order = 1;
warning('stochastic_solvers: using order = 1 because Hessian is equal to zero');
end
if options_.order>2 && ~options_.k_order_solver
error('You need to set k_order_solver for order>2')
end
if (options_.aim_solver == 1) && (local_order > 1)
error('Option "aim_solver" is incompatible with order >= 2')

6
matlab/utilities/dataset/makedataset.m
View File

@ -24,7 +24,7 @@ function [DynareDataset, DatasetInfo, newdatainterface] = makedataset(DynareOpti
%
% See also dynare_estimation_init
% Copyright (C) 2014-2017 Dynare Team
% Copyright (C) 2014-2018 Dynare Team
%
% This file is part of Dynare.
%
@ -81,7 +81,7 @@ elseif isempty(DynareOptions.datafile) && ~isempty(DynareOptions.dataset.series)
elseif ~isempty(DynareOptions.datafile) && isempty(DynareOptions.dataset.file)
datafile = DynareOptions.datafile;
newdatainterface = 0;
elseif isempty(DynareOptions.datafile) && ~isempty(DynareOptions.dataset.file)
elseif ~isempty(DynareOptions.datafile) && ~isempty(DynareOptions.dataset.file)
error('makedataset: You cannot simultaneously use the data command and the datafile option (in the estimation command)!')
else
error('makedataset: You have to specify the datafile!')
@ -267,7 +267,7 @@ else
end
% Compute the empirical mean of the observed variables.
DatasetInfo.descriptive.mean = nanmean(DynareDataset.data);
DatasetInfo.descriptive.mean = nanmean(DynareDataset.data,1);
% Compute the empirical covariance matrix of the observed variables.
DatasetInfo.descriptive.covariance = nancovariance(DynareDataset.data);

4
matlab/utilities/dataset/nancovariance.m
View File

@ -51,7 +51,7 @@ function CovarianceMatrix = nancovariance(data)
CovarianceMatrix = zeros(size(data,2));
if isanynan(data)
data = bsxfun(@minus,data,nanmean(data));
data = bsxfun(@minus,data,nanmean(data,1));
for i=1:size(data,2)
for j=i:size(data,2)
CovarianceMatrix(i,j) = nanmean(data(:,i).*data(:,j));
@ -61,7 +61,7 @@ if isanynan(data)
end
end
else
data = bsxfun(@minus,data,mean(data));
data = bsxfun(@minus,data,mean(data,1));
CovarianceMatrix = (transpose(data)*data)/size(data,1);
end

14
matlab/utilities/dataset/quarterly2annual.m
View File

@ -14,6 +14,7 @@ function [ya, yass, gya, gyass] = quarterly2annual(y,yss,GYTREND0,type,islog,aux
% 5 annual price as quantity weighted average
% 6 annual quantity from average price
% 7 annual nominal from Q real and deflator
% 8 annual ratio [e.g. trade balance to GDP]
% islog 0 level (default)
% 1 log-level
% 2 growth rate Q frequency
@ -120,6 +121,19 @@ switch type
yn = (y+yss).*(yaux+yauxss) - yss.*yauxss;
[ya, yass] = quarterly2annual(yn,yss.*yauxss,GYTREND0+GYTREND0aux,typeaux,0,0);
GYTREND0=GYTREND0+GYTREND0aux;
case 8
% numerator
yn = y;
[yna, ynass] = quarterly2annual(yn,yss,GYTREND0,typeaux(1),0,0);
% denominator
yd = yaux;
[yda, ydass] = quarterly2annual(yd,yauxss,GYTREND0aux,typeaux(2),0,0);
% ratio
yass = ynass/ydass;
ya = (yna+ynass)./(yda+ydass)-yass;
GYTREND0 = GYTREND0 - GYTREND0aux;
otherwise
error('Wrong type input')
end

47
matlab/utilities/general/demean.m
View File

@ -1,33 +1,14 @@
function c = demean(x)
% Removes the mean of each column of a matrix.
%
% INPUTS
% - x [double] T*N matrix of data.
%
% OUTPUTS
% - c [double] T*N matrix of demeaned data.
%@info:
%! @deftypefn {Function File} {@var{c} =} demean (@var{x})
%! @anchor{demean}
%! This function removes the mean of each column of a matrix.
%!
%! @strong{Inputs}
%! @table @var
%! @item x
%! Matlab matrix (T-by-N).
%! @end table
%!
%! @strong{Outputs}
%! @table @var
%! @item c
%! Matlab matrix (T-by-N). The demeaned x matrix.
%! @end table
%!
%! @strong{This function is called by:}
%! @ref{compute_cova}, @ref{compute_acov}, @ref{compute_std}.
%!
%! @strong{This function calls:}
%! @ref{ndim},
%!
%! @end deftypefn
%@eod:
% Copyright (C) 2011-2017 Dynare Team
% Copyright (C) 2011-2018 Dynare Team
%
% This file is part of Dynare.
%
@ -44,12 +25,8 @@ function c = demean(x)
% You should have received a copy of the GNU General Public License
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
% Original author: stephane DOT adjemian AT univ DASH lemans DOT fr
if ndim(x)>2
error('This function is not implemented for arrays with dimension greater than two!')
end
if ndim(x)==1
c = x-mean(x);
elseif ndim(x)==2
c = bsxfun(@minus,x,mean(x));
else
error('descriptive_statistics::demean:: This function is not implemented for arrays with dimension greater than two!')
end
c = bsxfun(@minus, x, nanmean(x));

2
preprocessor/ComputingTasks.cc
View File

@ -2686,7 +2686,7 @@ InitialConditionDecompositionStatement::writeOutput(ostream &output, const strin
output << "options_ = set_default_initial_condition_decomposition_options(options_);" << endl;
options_list.writeOutput(output);
symbol_list.writeOutput("var_list_", output);
output << "initial_condition_decomposition(M_, oo_, options_, var_list_, bayestopt_, estim_params_);" << endl;
output << "oo_ = initial_condition_decomposition(M_, oo_, options_, var_list_, bayestopt_, estim_params_);" << endl;
}
ConditionalForecastStatement::ConditionalForecastStatement(const OptionsList &options_list_arg) :

7
preprocessor/DynareBison.yy
View File

@ -114,7 +114,7 @@ class ParsingDriver;
%token CPF_WEIGHTS AMISANOTRISTANI MURRAYJONESPARSLOW WRITE_EQUATION_TAGS METHOD
%token NONLINEAR_FILTER_INITIALIZATION FILTER_ALGORITHM PROPOSAL_APPROXIMATION CUBATURE UNSCENTED MONTECARLO DISTRIBUTION_APPROXIMATION
%token <string_val> NAME
%token USE_PENALIZED_OBJECTIVE_FOR_HESSIAN INIT_STATE RESCALE_PREDICTION_ERROR_COVARIANCE GENERATE_IRFS
%token USE_PENALIZED_OBJECTIVE_FOR_HESSIAN INIT_STATE FAST_REALTIME RESCALE_PREDICTION_ERROR_COVARIANCE GENERATE_IRFS
%token NAN_CONSTANT NO_STATIC NOBS NOCONSTANT NODISPLAY NOCORR NODIAGNOSTIC NOFUNCTIONS NO_HOMOTOPY
%token NOGRAPH POSTERIOR_NOGRAPH POSTERIOR_GRAPH NOMOMENTS NOPRINT NORMAL_PDF SAVE_DRAWS MODEL_NAME STDERR_MULTIPLES DIAGONAL_ONLY
%token OBSERVATION_TRENDS OPTIM OPTIM_WEIGHTS ORDER OSR OSR_PARAMS MAX_DIM_COVA_GROUP ADVANCED OUTFILE OUTVARS OVERWRITE
@ -2827,6 +2827,7 @@ realtime_shock_decomposition_option : o_parameter_set
| o_shock_decomposition_presample
| o_shock_decomposition_forecast
| o_save_realtime
| o_fast_realtime
;
plot_shock_decomposition_options_list : plot_shock_decomposition_option COMMA plot_shock_decomposition_options_list
@ -2860,6 +2861,9 @@ initial_condition_decomposition_option : o_icd_type
| o_icd_write_xls
| o_icd_plot_init_date
| o_icd_plot_end_date
| o_nodisplay
| o_graph_format
| o_psd_fig_name
;
homotopy_setup: HOMOTOPY_SETUP ';' homotopy_list END ';'
@ -3240,6 +3244,7 @@ o_init_state : INIT_STATE EQUAL INT_NUMBER { driver.option_num("shock_decomp.ini
o_shock_decomposition_presample : PRESAMPLE EQUAL INT_NUMBER { driver.option_num("shock_decomp.presample", $3); };
o_shock_decomposition_forecast : FORECAST EQUAL INT_NUMBER { driver.option_num("shock_decomp.forecast", $3); };
o_save_realtime : SAVE_REALTIME EQUAL vec_int { driver.option_vec_int("shock_decomp.save_realtime", $3); };
o_fast_realtime : FAST_REALTIME EQUAL INT_NUMBER { driver.option_num("shock_decomp.fast_realtime", $3); };
o_nodisplay : NODISPLAY { driver.option_num("nodisplay","1"); };
o_psd_nodisplay : NODISPLAY { driver.option_num("plot_shock_decomp.nodisplay","1"); };
o_graph_format : GRAPH_FORMAT EQUAL allowed_graph_formats

1
preprocessor/DynareFlex.ll
View File

@ -594,6 +594,7 @@ DATE -?[0-9]+([YyAa]|[Mm]([1-9]|1[0-2])|[Qq][1-4]|[Ww]([1-9]{1}|[1-4][0-9]|5[0-2
<DYNARE_STATEMENT>controlled_varexo {return token::CONTROLLED_VAREXO; }
<DYNARE_STATEMENT>parameter_set {return token::PARAMETER_SET; }
<DYNARE_STATEMENT>init_state {return token::INIT_STATE; }
<DYNARE_STATEMENT>fast_realtime {return token::FAST_REALTIME; }
<DYNARE_STATEMENT>save_realtime {return token::SAVE_REALTIME;}
<DYNARE_STATEMENT>detail_plot {return token::DETAIL_PLOT;}
<DYNARE_STATEMENT>interactive {return token::INTERACTIVE;}

8
preprocessor/macro/MacroDriver.cc
View File

@ -1,5 +1,5 @@
/*
* Copyright (C) 2008-2017 Dynare Team
* Copyright (C) 2008-2018 Dynare Team
*
* This file is part of Dynare.
*
@ -60,7 +60,11 @@ MacroDriver::parse(const string &f, const string &fb, const string &modfiletxt,
}
catch (boost::bad_lexical_cast &)
{
file_with_endl << "@#define " << it->first << " = \"" << it->second << "\"" << endl;
if (!it->second.empty() && it->second.at(0) == '[' && it->second.at(it->second.length()-1) == ']')
// If the input is an array. Issue #1578
file_with_endl << "@#define " << it->first << " = " << it->second << endl;
else
file_with_endl << "@#define " << it->first << " = \"" << it->second << "\"" << endl;
}
file_with_endl << modfiletxt << endl;

14
preprocessor/macro/MacroFlex.ll
View File

@ -1,5 +1,5 @@
/*
* Copyright (C) 2008-2017 Dynare Team
* Copyright (C) 2008-2018 Dynare Team
*
* This file is part of Dynare.
*
@ -71,6 +71,8 @@ CONT \\\\
yylloc->step();
%}
<STMT,EXPR,FOR_BODY,THEN_BODY,ELSE_BODY>["/"]["/"].*
<INITIAL>^{SPC}*@#{SPC}*includepath{SPC}+\"([^\"\r\n:;|<>]*){1}(:[^\"\r\n:;|<>]*)*\"{SPC}*{EOL} {
yylloc->lines(1);
yylloc->step();
@ -303,6 +305,16 @@ CONT \\\\
then_body_tmp.append(yytext);
yylloc->step();
}
<THEN_BODY>^{SPC}*@#{SPC}*ifdef({SPC}|{CONT}) {
nested_if_nb++;
then_body_tmp.append(yytext);
yylloc->step();
}
<THEN_BODY>^{SPC}*@#{SPC}*ifndef({SPC}|{CONT}) {
nested_if_nb++;
then_body_tmp.append(yytext);
yylloc->step();
}
<THEN_BODY>. { then_body_tmp.append(yytext); yylloc->step(); }
<THEN_BODY><<EOF>> { driver.error(if_stmt_loc_tmp, "@#if/@#ifdef/@#ifndef not matched by an @#endif or file does not end with a new line (unexpected end of file)"); }
<THEN_BODY>^{SPC}*@#{SPC}*else{SPC}*(\/\/.*)?{EOL} {

6
tests/shock_decomposition/ls2003_plot.mod
View File

@ -129,6 +129,12 @@ close all
plot_shock_decomposition(detail_plot, interactive, use_shock_groups = row, type = qoq);
plot_shock_decomposition(detail_plot, interactive, realtime = 3, vintage = 29);
close all,
// testing realtime decomposition with fast_realtime option
realtime_shock_decomposition(fast_realtime=75);
collect_latex_files;
if system(['pdflatex -halt-on-error -interaction=batchmode ' M_.fname '_TeX_binder.tex'])
error('TeX-File did not compile.')

2
windows/README.txt
View File

@ -61,7 +61,7 @@ This version of Dynare is compiled for Octave 4.2.1 (MinGW, 32bit and 64bit), an
with other versions of Octave. The recommended version of Octave can be
downloaded at:
http://www.dynare.org/download/octave/windows
https://www.gnu.org/software/octave/download.html
Every time you run Octave, you should type the two following commands (assuming
that you have installed Dynare at the standard location, and replacing '4.x.y'