Added integration test for nonlinear filters with unit roots.
Stéphane Adjemian (Charybdis)
parent
4dcdb81a31
commit
c1434118bd
|
|
@ -338,6 +338,10 @@ MODFILES = \
|
|||
observation_trends_and_prefiltering/calib_smoother/Tr_prefil_f_obs_loglin_cal_smoother.mod \
|
||||
observation_trends_and_prefiltering/ML/Trend_no_prefilter_selected_var.mod
|
||||
|
||||
PARTICLEFILES = \
|
||||
particle/dsge_base2.mod \
|
||||
particle/dsge_unit_root.mod
|
||||
|
||||
|
||||
XFAIL_MODFILES = ramst_xfail.mod \
|
||||
estim_param_in_shock_value_xfail.mod \
|
||||
|
|
@ -694,6 +698,10 @@ gradient: m/gradient o/gradient
|
|||
m/gradient: $(patsubst %.mod, %.m.trs, $(filter gradient/%.mod, $(MODFILES)))
|
||||
o/gradient: $(patsubst %.mod, %.o.trs, $(filter gradient/%.mod, $(MODFILES)))
|
||||
|
||||
particle: m/particle o/particle
|
||||
m/particle: $(patsubst %.mod, %.m.trs, $(PARTICLEFILES))
|
||||
o/particle: $(patsubst %.mod, %.o.trs, $(PARTICLEFILES))
|
||||
|
||||
# Matlab TRS Files
|
||||
M_TRS_FILES = $(patsubst %.mod, %.m.trs, $(MODFILES))
|
||||
M_TRS_FILES += run_block_byte_tests_matlab.m.trs run_reporting_test_matlab.m.trs run_all_unitary_tests.m.trs
|
||||
|
|
@ -710,8 +718,6 @@ M_TLS_FILES = $(patsubst %.m, %.m.tls, $(MFILES))
|
|||
# Octave TLS Files
|
||||
O_TLS_FILES = $(patsubst %.m, %.o.tls, $(MFILES))
|
||||
|
||||
|
||||
|
||||
EXTRA_DIST = \
|
||||
read_trs_files.sh \
|
||||
run_test_matlab.m \
|
||||
|
|
|
|||
|
|
@ -1,27 +1,51 @@
|
|||
// This file deals with the resolution and estimation of a basic DSGE model with
|
||||
//employment for comparison with the benchmark in Gauss which solves with
|
||||
//the same particular filter but global methodology.
|
||||
//
|
||||
// Juin 2015
|
||||
|
||||
// DGP
|
||||
@#ifndef RISKY_CALIBRATION
|
||||
@#define RISKY_CALIBRATION = 0
|
||||
@#endif
|
||||
@#ifndef EXTREME_CALIBRATION
|
||||
@#define EXTREME_CALIBRATION = 1
|
||||
@#endif
|
||||
@#ifndef BENCHMARK_CALIBRATION
|
||||
@#define BENCHMARK_CALIBRATION = 0
|
||||
@#endif
|
||||
|
||||
// ALGORITHM
|
||||
@#define LINEAR_KALMAN = 1
|
||||
@#ifndef LINEAR_KALMAN
|
||||
@#define LINEAR_KALMAN = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_SIR
|
||||
@#define ALGO_SIR = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_SISmoothR
|
||||
@#define ALGO_SISmoothR = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_APF
|
||||
@#define ALGO_APF = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_GPF
|
||||
@#define ALGO_GPF = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_GCF
|
||||
@#define ALGO_GCF = 0
|
||||
@#define ALGO_GUF = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_GUF
|
||||
@#define ALGO_GUF = 1
|
||||
@#endif
|
||||
@#ifndef ALGO_GMPF
|
||||
@#define ALGO_GMPF = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_GMCF
|
||||
@#define ALGO_GMCF = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_ONLINE_1
|
||||
@#define ALGO_ONLINE_1 = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_ONLINE_2
|
||||
@#define ALGO_ONLINE_2 = 0
|
||||
|
||||
@#endif
|
||||
@#ifndef MCMC
|
||||
@#define MCMC = 0
|
||||
@#endif
|
||||
|
||||
var k A c l i y;
|
||||
varexo e_a;
|
||||
|
|
@ -164,7 +188,9 @@ options_.mode_check.number_of_points = 250;
|
|||
estimation(order=1,number_of_particles=1000,mode_compute=11);
|
||||
@#endif
|
||||
|
||||
options_.mh_nblck = 10 ;
|
||||
options_.posterior_sampling_method = 'RWGMH';
|
||||
options_.rwgmh_scale_shock = (1e-5)*[10 10 1 1 10 10 10 1000 10 10] ;
|
||||
estimation(order=1,mh_replic=5000,mode_compute=0,mode_file=dsge_base2_mode);
|
||||
@#if MCMC
|
||||
options_.mh_nblck = 10 ;
|
||||
options_.posterior_sampling_method = 'RWGMH';
|
||||
options_.rwgmh_scale_shock = (1e-5)*[10 10 1 1 10 10 10 1000 10 10] ;
|
||||
estimation(order=1,mh_replic=5000,mode_compute=0,mode_file=dsge_base2_mode);
|
||||
@#endif
|
||||
|
|
@ -0,0 +1,221 @@
|
|||
// DGP
|
||||
@#ifndef RISKY_CALIBRATION
|
||||
@#define RISKY_CALIBRATION = 0
|
||||
@#endif
|
||||
@#ifndef EXTREME_CALIBRATION
|
||||
@#define EXTREME_CALIBRATION = 1
|
||||
@#endif
|
||||
@#ifndef BENCHMARK_CALIBRATION
|
||||
@#define BENCHMARK_CALIBRATION = 0
|
||||
@#endif
|
||||
|
||||
// ALGORITHM
|
||||
@#ifndef LINEAR_KALMAN
|
||||
@#define LINEAR_KALMAN = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_SIR
|
||||
@#define ALGO_SIR = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_SISmoothR
|
||||
@#define ALGO_SISmoothR = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_APF
|
||||
@#define ALGO_APF = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_GPF
|
||||
@#define ALGO_GPF = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_GCF
|
||||
@#define ALGO_GCF = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_GUF
|
||||
@#define ALGO_GUF = 1
|
||||
@#endif
|
||||
@#ifndef ALGO_GMPF
|
||||
@#define ALGO_GMPF = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_GMCF
|
||||
@#define ALGO_GMCF = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_ONLINE_1
|
||||
@#define ALGO_ONLINE_1 = 0
|
||||
@#endif
|
||||
@#ifndef ALGO_ONLINE_2
|
||||
@#define ALGO_ONLINE_2 = 0
|
||||
@#endif
|
||||
@#ifndef MCMC
|
||||
@#define MCMC = 0
|
||||
@#endif
|
||||
|
||||
var k A c l i y;
|
||||
varexo e_a;
|
||||
|
||||
parameters alp bet tet tau delt rho ;
|
||||
alp = 0.4;
|
||||
bet = 0.99;
|
||||
tet = 0.357 ;
|
||||
tau = 50 ;
|
||||
delt = 0.02;
|
||||
rho = 1.0;
|
||||
|
||||
model;
|
||||
c = ((1 - alp)*tet/(1-tet))*A*(1-l)*((k(-1)/l)^alp) ;
|
||||
y = A*(k(-1)^alp)*(l^(1-alp)) ;
|
||||
i = y-c ;
|
||||
k = (1-delt)*k(-1) + i ;
|
||||
log(A) = rho*log(A(-1)) + e_a ;
|
||||
(((c^(tet))*((1-l)^(1-tet)))^(1-tau))/c - bet*((((c(+1)^(tet))*((1-l(+1))^(1-tet)))^(1-tau))/c(+1))*(1 -delt+alp*(A(1)*(k^alp)*(l(1)^(1-alp)))/k)=0 ;
|
||||
end;
|
||||
|
||||
shocks;
|
||||
var e_a; stderr 0.035;
|
||||
end;
|
||||
|
||||
steady(nocheck);
|
||||
|
||||
|
||||
stoch_simul(periods=5000, irf=0, noprint);
|
||||
|
||||
@#if EXTREME_CALIBRATION
|
||||
verbatim;
|
||||
y = y + randn(size(y))*.0175;
|
||||
l = l + randn(size(l))*.00312;
|
||||
i = i + randn(size(i))*.00465;
|
||||
end;
|
||||
ds = dseries([y, l, i, c, A], 1900Q1, {'y'; 'l'; 'i'; 'c'; 'A'});
|
||||
@#endif
|
||||
|
||||
@#if RISKY_CALIBRATION
|
||||
verbatim;
|
||||
y = y + randn(size(y))*.00158;
|
||||
l = l + randn(size(l))*.0011;
|
||||
i = i + randn(size(i))*.000866;
|
||||
end;
|
||||
ds = dseries([y, l, i, c, A], 1900Q1, {'y'; 'l'; 'i'; 'c'; 'A'});
|
||||
@#endif
|
||||
|
||||
@#if BENCHMARK_CALIBRATION
|
||||
// TODO: This section is to be completed (measurement are missing)
|
||||
ds = dseries([y, l, i, c, A], 1900Q1, {'y'; 'l'; 'i'; 'c'; 'A'});
|
||||
@#endif
|
||||
|
||||
estimated_params;
|
||||
alp, uniform_pdf,,, 0.0001, 0.99;
|
||||
bet, uniform_pdf,,, 0.0001, 0.99999;
|
||||
tet, uniform_pdf,,, 0.0001, .999;
|
||||
tau, uniform_pdf,,, 0.0001, 100;
|
||||
delt, uniform_pdf,,, 0.0001, 0.05;
|
||||
stderr e_a, uniform_pdf,,, 0.00001, 0.1;
|
||||
stderr y, uniform_pdf,,, 0.00001, 0.1;
|
||||
stderr l, uniform_pdf,,, 0.00001, 0.1;
|
||||
stderr i, uniform_pdf,,, 0.00001, 0.1;
|
||||
end;
|
||||
|
||||
@#if RISKY_CALIBRATION
|
||||
estimated_params_init;
|
||||
alp, 0.4;
|
||||
bet, 0.99;
|
||||
tet, 0.357;
|
||||
tau, 50;
|
||||
delt, 0.02;
|
||||
stderr e_a, .035;
|
||||
stderr y, .00158;
|
||||
stderr l, .0011;
|
||||
stderr i, .000866;
|
||||
end;
|
||||
@#endif
|
||||
|
||||
@#if EXTREME_CALIBRATION
|
||||
estimated_params_init;
|
||||
alp, 0.4;
|
||||
bet, 0.99;
|
||||
tet, 0.357;
|
||||
tau, 50;
|
||||
delt, 0.02;
|
||||
stderr e_a, .035;
|
||||
stderr y, .0175;
|
||||
stderr l, .00312;
|
||||
stderr i, .00465;
|
||||
end;
|
||||
@#endif
|
||||
|
||||
varobs y l i ;
|
||||
|
||||
options_.mode_check.neighbourhood_size = .01 ;
|
||||
options_.mode_check.number_of_points = 250;
|
||||
|
||||
//options_.particle.initialization = 3;
|
||||
|
||||
@#if EXTREME_CALIBRATION
|
||||
data(series=ds, nobs=200, first_obs=1973Q1);
|
||||
@#endif
|
||||
|
||||
@#if RISKY_CALIBRATION
|
||||
data(series=ds, nobs=200, first_obs=1973Q1);
|
||||
@#endif
|
||||
|
||||
@#if BENCHMARK_CALIBRATION
|
||||
data(series=ds, nobs = 200, first_obs=1973Q1);
|
||||
@#endif
|
||||
|
||||
|
||||
@#if LINEAR_KALMAN
|
||||
% estimation(nograph,order=1,mode_compute=8,mh_replic=0,mode_check);
|
||||
@#endif
|
||||
|
||||
@#if ALGO_SIR
|
||||
estimation(order=2,nonlinear_filter_initialization=3,nograph,number_of_particles=1000,mh_replic=0,mode_compute=8,mode_check);
|
||||
@#endif
|
||||
|
||||
@#if ALGO_SISmoothR
|
||||
estimation(order=2,nonlinear_filter_initialization=3,nograph,number_of_particles=1000,resampling_method=smooth,mode_compute=8,mh_replic=0);
|
||||
estimation(order=2,nonlinear_filter_initialization=3,nograph,number_of_particles=1000,resampling_method=smooth,mode_compute=8,mode_file=dsge_unit_root_mode,mh_replic=0);
|
||||
estimation(order=2,nonlinear_filter_initialization=3,nograph,number_of_particles=1000,resampling_method=smooth,mode_compute=4,mode_file=dsge_unit_root_mode,mh_replic=0,mode_check);
|
||||
@#endif
|
||||
|
||||
@#if ALGO_APF
|
||||
estimation(order=2,nonlinear_filter_initialization=3,nograph,filter_algorithm=apf,number_of_particles=1000,mh_replic=0,mode_compute=8,mode_check);
|
||||
@#endif
|
||||
|
||||
@#if ALGO_GPF
|
||||
estimation(order=2,nonlinear_filter_initialization=3,nograph,filter_algorithm=gf,distribution_approximation=montecarlo,number_of_particles=1000,mh_replic=0,mode_compute=8);
|
||||
estimation(order=2,nonlinear_filter_initialization=3,nograph,filter_algorithm=gf,distribution_approximation=montecarlo,number_of_particles=1000,mode_file=dsge_unit_root_mode,mh_replic=0,mode_compute=4,mode_check);
|
||||
@#endif
|
||||
|
||||
@#if ALGO_GCF
|
||||
estimation(order=2,nonlinear_filter_initialization=3,nograph,filter_algorithm=gf,mh_replic=0,mode_compute=8);
|
||||
estimation(order=2,nonlinear_filter_initialization=3,nograph,filter_algorithm=gf,mh_replic=0,mode_compute=4,mode_file=dsge_unit_root_mode,mode_check);
|
||||
@#endif
|
||||
|
||||
@#if ALGO_GUF
|
||||
estimation(order=2,nonlinear_filter_initialization=3,nograph,filter_algorithm=gf,proposal_approximation=unscented,distribution_approximation=unscented,mh_replic=0,mode_compute=4);
|
||||
estimation(order=2,nonlinear_filter_initialization=3,nograph,filter_algorithm=gf,proposal_approximation=unscented,distribution_approximation=unscented,mode_file=dsge_unit_root_mode,mh_replic=0,mode_compute=8,mode_check);
|
||||
@#endif
|
||||
|
||||
@#if ALGO_GMPF
|
||||
estimation(nograph,order=2,nonlinear_filter_initialization=3,filter_algorithm=gmf,distribution_approximation=montecarlo,number_of_particles=1000,mh_replic=0,mode_compute=8);
|
||||
estimation(nograph,order=2,nonlinear_filter_initialization=3,filter_algorithm=gmf,distribution_approximation=montecarlo,number_of_particles=1000,mh_replic=0,mode_file=dsge_unit_root_mode,mode_compute=8);
|
||||
estimation(nograph,order=2,nonlinear_filter_initialization=3,filter_algorithm=gmf,distribution_approximation=montecarlo,number_of_particles=1000,mh_replic=0,mode_file=dsge_unit_root_mode,mode_compute=4,mode_check);
|
||||
@#endif
|
||||
|
||||
@#if ALGO_GMCF
|
||||
estimation(nograph,order=2,nonlinear_filter_initialization=3,filter_algorithm=gmf,mh_replic=0,mode_compute=8);
|
||||
estimation(nograph,order=2,nonlinear_filter_initialization=3,filter_algorithm=gmf,mh_replic=0,mode_compute=4,mode_file=dsge_unit_root_mode,mode_check);
|
||||
@#endif
|
||||
|
||||
@#if ALGO_ONLINE_2
|
||||
options_.particle.liu_west_delta = 0.9 ;
|
||||
estimation(order=2,number_of_particles=1000,mode_compute=11);
|
||||
@#endif
|
||||
|
||||
@#if ALGO_ONLINE_1
|
||||
options_.particle.liu_west_delta = 0.9 ;
|
||||
estimation(order=1,number_of_particles=1000,mode_compute=11);
|
||||
@#endif
|
||||
|
||||
@#if MCMC
|
||||
options_.mh_nblck = 10 ;
|
||||
options_.posterior_sampling_method = 'RWGMH';
|
||||
options_.rwgmh_scale_shock = (1e-5)*[10 10 1 1 10 10 10 1000 10 10] ;
|
||||
estimation(order=1,mh_replic=5000,mode_compute=0,mode_file=dsge_base2_mode);
|
||||
@#endif
|
||||
Loading…
Reference in New Issue