Create on the fly a routine for evaluating the (VAR/PAC) expectations.

The routine takes a dseries object as unique argument and return an updated
object with the expectation term.

If the mod file is named `example.mod` and if the (VAR/PAC) expectation model is
named `toto`, then after

var_expectation.print('toto');

the expectation term can be evaluated:

ts = example.var_expectations.evaluate_varexp(ts);

where ts is a dseries object containing all the time series appearign in the
auxiliary (var or trend_component).
time-shift
Stéphane Adjemia (Scylla) 2018-12-03 15:07:43 +01:00
parent 53ab321de8
commit cb4384bb27
Signed by untrusted user who does not match committer: stepan
GPG Key ID: A6D44CB9C64CE77B
3 changed files with 187 additions and 3 deletions

View File

@ -92,8 +92,9 @@ end
auxmodel = M_.(expectationmodel.auxiliary_model_type).(expectationmodel.auxiliary_model_name);
%% First print the list of parameters appearing in the VAR_EXPECTATION/PAC_EXPECTATION term.
%
% First print the list of parameters appearing in the VAR_EXPECTATION/PAC_EXPECTATION term.
%
if ~exist(sprintf('%s/model/%s', M_.fname, expectationmodelkind), 'dir')
mkdir(sprintf('%s/model/%s', M_.fname, expectationmodelkind))
end
@ -154,7 +155,9 @@ end
fclose(fid);
%% Second print the expanded VAR_EXPECTATION/PAC_EXPECTATION term.
%
% Second print the expanded VAR_EXPECTATION/PAC_EXPECTATION term.
%
filename = sprintf('%s/model/%s/%s-expression.inc', M_.fname, expectationmodelkind, expectationmodelname);
fid = fopen(filename, 'w');
fprintf(fid, '// This file has been generated by dynare (%s).\n', datestr(now));
@ -238,4 +241,115 @@ for i=1:maxlag
end
fprintf(fid, '%s', expression);
fclose(fid);
%
% Third print a routine for evaluating VAR_EXPECTATION/PAC_EXPECTATION term (updates a dseries object).
%
kind = strrep(expectationmodelkind, '-', '_');
mkdir(sprintf('+%s/+%s', M_.fname, kind));
filename = sprintf('+%s/+%s/evaluate_%s.m', M_.fname, kind, expectationmodelname);
fid = fopen(filename, 'w');
fprintf(fid, 'function dbase = evaluate_%s(dbase)\n\n', expectationmodelname);
fprintf(fid, '%% Evaluates %s term (%s), updating a dseries object.\n', kind, expectationmodelname);
fprintf(fid, '%%\n');
fprintf(fid, '%% INPUTS\n');
fprintf(fid, '%% - dbase [dseries] databse containing all the variables appearing in the auxiliary model for the expectation.\n');
fprintf(fid, '%%\n');
fprintf(fid, '%% OUTPUTS\n');
fprintf(fid, '%% - dbase [dseries] same databse augmented with the expectation term .\n');
fprintf(fid, '%%\n');
fprintf(fid, '%% REMARKS\n');
fprintf(fid, '%% The name of the appended variable in dbase is the declared name for the (PAC/VAR) expectation model.\n\n');
fprintf(fid, '%% This file has been generated by dynare (%s).\n\n', datestr(now));
id = 0;
maxlag = max(auxmodel.max_lag);
if isequal(expectationmodel.auxiliary_model_type, 'trend_component')
% Need to add a lag since the error correction equations are rewritten in levels.
maxlag = maxlag+1;
end
for i=1:maxlag
for j=1:length(auxmodel.list_of_variables_in_companion_var)
id = id+1;
variable = auxmodel.list_of_variables_in_companion_var{j};
transformations = {};
ida = get_aux_variable_id(variable);
while ida
if isequal(M_.aux_vars(ida).type, 8)
transformations = [transformations, 'diff'];
variable = M_.endo_names{M_.aux_vars(ida).orig_index};
ida = get_aux_variable_id(variable);
elseif isequal(M_.aux_vars(ida).type, 10)
transformations = [transformations, 'log'];
variable = M_.endo_names{M_.aux_vars(ida).orig_index};
ida = get_aux_variable_id(variable);
end
end
switch expectationmodelkind
case 'var-expectations'
parameter = M_.params(expectationmodel.param_indices(id));
case 'pac-expectations'
parameter = 0;
if isfield(expectationmodel,'h0_param_indices') && ~isempty(expectationmodel.h0_param_indices)
parameter = M_.params(expectationmodel.h0_param_indices(id));
end
if isfield(expectationmodel,'h1_param_indices') && ~isempty(expectationmodel.h1_param_indices)
if ~parameter
parameter = M_.params(expectationmodel.h1_param_indices(id));
else
parameter = parameter+M_.params(expectationmodel.h1_param_indices(id));
end
end
otherwise
end
switch expectationmodelkind
case 'var-expectations'
if i>1
variable = sprintf('dbase.%s(-%d)', variable, i-1);
else
variable = sprintf('dbase.%s', variable);
end
case 'pac-expectations'
variable = sprintf('dbase.%s(-%d)', variable, i);
otherwise
end
if ~isempty(transformations)
for k=length(transformations):-1:1
variable = sprintf('%s.%s()', variable, transformations{k});
end
end
if isequal(id, 1)
if isequal(expectationmodelkind, 'pac-expectations') && growth_correction
pgrowth = M_.param_names(expectationmodel.growth_neutrality_param_index);
switch expectationmodel.growth_type
case 'parameter'
vgrowth = M_.param_names{expectationmodel.growth_index};
case 'endogenous'
vgrowth = M_.endo_names{expectationmodel.growth_index};
case 'exogenous'
vgrowth = M_.exo_names{expectationmodel.growth_index};
otherwise
end
if parameter>=0
expression = sprintf('%s*%s+%s*%s', num2str(pgrowth, '%1.16f'), vgrowth, num2str(parameter, '%1.16f'), variable);
else
expression = sprintf('%s*%s-%s*%s', num2str(pgrowth, '%1.16f'), vgrowth, num2str(-parameter, '%1.16f'), variable);
end
else
expression = sprintf('%s*%s', num2str(parameter, '%1.16f'), variable);
end
else
if parameter>=0
expression = sprintf('%s + %s*%s', expression, num2str(parameter, '%1.16f'), variable);
else
expression = sprintf('%s - %s*%s', expression, num2str(-parameter, '%1.16f'), variable);
end
end
end
end
fprintf(fid, 'dbase.%s = %s;', expectationmodelname, expression);
fclose(fid);

View File

@ -369,6 +369,7 @@ MODFILES = \
var-expectations/8/example.mod \
var-expectations/8/substitution.mod \
var-expectations/9/example.mod \
var-expectations/10/example.mod \
trend-component-and-var-models/vm1.mod \
trend-component-and-var-models/vm2.mod \
trend-component-and-var-models/vm3.mod \

View File

@ -0,0 +1,69 @@
// --+ options: stochastic,json=compute +--
var foo z x y;
varexo e_x e_y e_z;
parameters a b c d e f beta ;
a = .9;
b = -.2;
c = .3;
f = .8;
d = .5;
e = .4;
beta = 1/(1+.02);
// Define a VAR model from a subset of equations in the model block.
var_model(model_name = toto, eqtags = [ 'X' 'Y' 'Z' ]);
// Define a VAR_EXPECTATION_MODEL
var_expectation_model(model_name = varexp, expression = diff(log(x)), auxiliary_model_name = toto, horizon = 1, discount = beta) ;
model;
[ name = 'X' ]
diff(log(x)) = a*diff(log(x(-1))) + b*diff(log(x(-2))) + c*diff(z(-2)) + e_x;
[ name = 'Z' ]
diff(z) = f*diff(z(-1)) + e_z;
[ name = 'Y' ]
log(y) = d*log(y(-2)) + e*diff(z(-1)) + e_y;
foo = var_expectation(varexp);
end;
// Initialize the VAR expectation model, will build the companion matrix of the VAR.
var_expectation.initialize('varexp')
// Update VAR_EXPECTATION reduced form parameters
var_expectation.update('varexp');
// Print expanded VAR_EXPECTATION expression in a file (to be included in substitution.mod).
var_expectation.print('varexp');
shocks;
var e_x = .01;
var e_y = .01;
var e_z = .01;
end;
verbatim;
initialconditions =zeros(3,4);
initialconditions(3,1) = .1; % foo(-1)
initialconditions(:,2) = .2; % y(-1)
initialconditions(3,3) = .3; % z(-1)
initialconditions(2,3) = .4; % z(-2)
initialconditions(3,4) = .5; % x(-1)
initialconditions(2,4) = .6; % x(-2)
initialconditions(1,4) = .7; % x(-3)
initialconditions = ...
dseries(initialconditions, dates('2000Q1'), {'foo', 'y','z', 'x'});
set_dynare_seed('default');
ts = simul_backward_model(initialconditions, 15);
foo = ts.foo.data;
% Evaluate the (VAR) expectation term
ts = example.var_expectations.evaluate_varexp(ts);
% Check tthat the evaluation is correct.
range = dates('2000Q4'):dates('2004Q2');
if max(abs(ts(range).foo.data-ts(range).varexp.data))>1e-5
error('Expectation term evaluations do not match!')
end
end;