Merge branch 'master' into ecb-master
commit
7d373f0d7e
|
@ -60,6 +60,7 @@ checksum
|
||||||
/doc/dynare.info
|
/doc/dynare.info
|
||||||
/doc/dynare.info-1
|
/doc/dynare.info-1
|
||||||
/doc/dynare.info-2
|
/doc/dynare.info-2
|
||||||
|
/doc/dynare.info-3
|
||||||
/doc/dynare.cp
|
/doc/dynare.cp
|
||||||
/doc/dynare.fn
|
/doc/dynare.fn
|
||||||
/doc/dynare.fns
|
/doc/dynare.fns
|
||||||
|
@ -212,3 +213,6 @@ tests/julia/rbc/rbc*.jl
|
||||||
|
|
||||||
# Octave variables saved when Octave crashes
|
# Octave variables saved when Octave crashes
|
||||||
octave-workspace
|
octave-workspace
|
||||||
|
|
||||||
|
# VERSION generated file
|
||||||
|
VERSION
|
805
NEWS
805
NEWS
|
@ -1,3 +1,800 @@
|
||||||
|
Announcement for Dynare 4.5.0 (on 2013-12-16)
|
||||||
|
=============================================
|
||||||
|
|
||||||
|
We are pleased to announce the release of Dynare 4.5.0.
|
||||||
|
|
||||||
|
This major release adds new features and fixes various bugs.
|
||||||
|
|
||||||
|
The Windows packages are already available for download at:
|
||||||
|
|
||||||
|
http://www.dynare.org/download/dynare-stable
|
||||||
|
|
||||||
|
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
|
||||||
|
9.2 (R2017a) and with GNU Octave version 4.2.
|
||||||
|
|
||||||
|
Here is the list of major user-visible changes:
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
Dynare 4.5
|
||||||
|
==========
|
||||||
|
|
||||||
|
|
||||||
|
- Ramsey policy
|
||||||
|
|
||||||
|
+ Added command `ramsey_model` that builds the expanded model with
|
||||||
|
FOC conditions for the planner's problem but doesn't perform any
|
||||||
|
computation. Usefull to compute Ramsey policy in a perfect
|
||||||
|
foresight model,
|
||||||
|
|
||||||
|
+ `ramsey_policy` accepts multipliers in its variable list and
|
||||||
|
displays results for them.
|
||||||
|
|
||||||
|
|
||||||
|
- Perfect foresight models
|
||||||
|
|
||||||
|
+ New commands `perfect_foresight_setup` (for preparing the
|
||||||
|
simulation) and `perfect_foresight_solver` (for computing it). The
|
||||||
|
old `simul` command still exist and is now an alias for
|
||||||
|
`perfect_foresight_setup` + `perfect_foresight_solver`. It is no
|
||||||
|
longer possible to manipulate by hand the contents of
|
||||||
|
`oo_.exo_simul` when using `simul`. People who want to do
|
||||||
|
it must first call `perfect_foresight_setup`, then do the
|
||||||
|
manipulations, then call `perfect_foresight_solver`,
|
||||||
|
|
||||||
|
+ By default, the perfect foresight solver will try a homotopy
|
||||||
|
method if it fails to converge at the first try. The old behavior
|
||||||
|
can be restored with the `no_homotopy` option,
|
||||||
|
|
||||||
|
+ New option `stack_solve_algo=7` that allows specifying a
|
||||||
|
`solve_algo` solver for solving the model,
|
||||||
|
|
||||||
|
+ New option `solve_algo` that allows specifying a solver for
|
||||||
|
solving the model when using `stack_solve_algo=7`,
|
||||||
|
|
||||||
|
+ New option `lmmcp` that solves the model via a Levenberg-Marquardt
|
||||||
|
mixed complementarity problem (LMMCP) solver,
|
||||||
|
|
||||||
|
+ New option `robust_lin_solve` that triggers the use of a robust
|
||||||
|
linear solver for the default `solve_algo=4`,
|
||||||
|
|
||||||
|
+ New options `tolf` and `tolx` to control termination criteria of
|
||||||
|
solvers,
|
||||||
|
|
||||||
|
+ New option `endogenous_terminal_period` to `simul`,
|
||||||
|
|
||||||
|
+ Added the possibility to set the initial condition of the
|
||||||
|
(stochastic) extended path simulations with the histval block.
|
||||||
|
|
||||||
|
|
||||||
|
- Optimal simple rules
|
||||||
|
|
||||||
|
+ Saves the optimal value of parameters to `oo_.osr.optim_params`,
|
||||||
|
|
||||||
|
+ New block `osr_params_bounds` allows specifying bounds for the
|
||||||
|
estimated parameters,
|
||||||
|
|
||||||
|
+ New option `opt_algo` allows selecting different optimizers while
|
||||||
|
the new option `optim` allows specifying the optimizer options,
|
||||||
|
|
||||||
|
+ The `osr` command now saves the names, bounds, and indices for the
|
||||||
|
estimated parameters as well as the indices and weights of the
|
||||||
|
variables entering the objective function into `M_.osr`.
|
||||||
|
|
||||||
|
|
||||||
|
- Forecasts and Smoothing
|
||||||
|
|
||||||
|
+ The smoother and forecasts take uncertainty about trends and means
|
||||||
|
into account,
|
||||||
|
|
||||||
|
+ Forecasts accounting for measurement error are now saved in fields
|
||||||
|
of the form `HPDinf_ME` and `HPDsup_ME`,
|
||||||
|
|
||||||
|
+ New fields `oo_.Smoother.Trend` and `oo_.Smoother.Constant` that
|
||||||
|
save the trend and constant parts of the smoothed variables,
|
||||||
|
|
||||||
|
+ new field `oo_.Smoother.TrendCoeffs` that stores the trend
|
||||||
|
coefficients.
|
||||||
|
|
||||||
|
+ Rolling window forecasts allowed in `estimation` command by
|
||||||
|
passing a vector to `first_obs`,
|
||||||
|
|
||||||
|
+ The `calib_smoother` command now accepts the `loglinear`,
|
||||||
|
`prefilter`, `first_obs` and `filter_decomposition` options.
|
||||||
|
|
||||||
|
|
||||||
|
- Estimation
|
||||||
|
|
||||||
|
+ New options: `logdata`, `consider_all_endogenous`,
|
||||||
|
`consider_only_observed`, `posterior_max_subsample_draws`,
|
||||||
|
`mh_conf_sig`, `diffuse_kalman_tol`, `dirname`, `nodecomposition`
|
||||||
|
|
||||||
|
+ `load_mh_file` and `mh_recover` now try to load chain's proposal density,
|
||||||
|
|
||||||
|
+ New option `load_results_after_load_mh` that allows loading some
|
||||||
|
posterior results from a previous run if no new MCMC draws are
|
||||||
|
added,
|
||||||
|
|
||||||
|
+ New option `posterior_nograph` that suppresses the generation of
|
||||||
|
graphs associated with Bayesian IRFs, posterior smoothed objects,
|
||||||
|
and posterior forecasts,
|
||||||
|
|
||||||
|
+ Saves the posterior density at the mode in
|
||||||
|
`oo_.posterior.optimization.log_density`,
|
||||||
|
|
||||||
|
+ The `filter_covariance` option now also works with posterior
|
||||||
|
sampling like Metropolis-Hastings,
|
||||||
|
|
||||||
|
+ New option `no_posterior_kernel_density` to suppress computation
|
||||||
|
of kernel density of posterior objects,
|
||||||
|
|
||||||
|
+ Recursive estimation and forecasting now provides the individual
|
||||||
|
`oo_` structures for each sample in `oo_recursive_`,
|
||||||
|
|
||||||
|
+ The `trace_plot` command can now plot the posterior density,
|
||||||
|
|
||||||
|
+ New command `generate_trace_plots` allows generating all trace
|
||||||
|
plots for one chain,
|
||||||
|
|
||||||
|
+ New commands `prior_function` and `posterior_function` that
|
||||||
|
execute a user-defined function on parameter draws from the
|
||||||
|
prior/posterior distribution,
|
||||||
|
|
||||||
|
+ New option `huge_number` for replacement of infinite bounds with
|
||||||
|
large number during `mode_compute`,
|
||||||
|
|
||||||
|
+ New option `posterior_sampling_method` allows selecting the new
|
||||||
|
posterior sampling options:
|
||||||
|
`tailored_random_block_metropolis_hastings` (Tailored randomized
|
||||||
|
block (TaRB) Metropolis-Hastings), `slice` (Slice sampler),
|
||||||
|
`independent_metropolis_hastings` (Independent
|
||||||
|
Metropolis-Hastings),
|
||||||
|
|
||||||
|
+ New option `posterior_sampler_options` that allow controlling the
|
||||||
|
options of the `posterior_sampling_method`, its `scale_file`-option
|
||||||
|
pair allows loading the `_mh_scale.mat`-file storing the tuned
|
||||||
|
scale factor from a previous run of `mode_compute=6`,
|
||||||
|
|
||||||
|
+ New option `raftery_lewis_diagnostics` that computes Raftery/Lewis
|
||||||
|
(1992) convergence diagnostics,
|
||||||
|
|
||||||
|
+ New option `fast_kalman_filter` that provides fast Kalman filter
|
||||||
|
using Chandrasekhar recursions as described in Ed Herbst (2015),
|
||||||
|
|
||||||
|
+ The `dsge_var` option now saves results at the posterior mode into
|
||||||
|
`oo_.dsge_var`,
|
||||||
|
|
||||||
|
+ New option `smoothed_state_uncertainty` to provide the uncertainty
|
||||||
|
estimate for the smoothed state estimate from the Kalman smoother,
|
||||||
|
|
||||||
|
+ New prior density: generalized Weibull distribution,
|
||||||
|
|
||||||
|
+ Option `mh_recover` now allows continuing a crashed chain at the
|
||||||
|
last save mh-file,
|
||||||
|
|
||||||
|
+ New option `nonlinear_filter_initialization` for the
|
||||||
|
`estimation` command. Controls the initial covariance matrix
|
||||||
|
of the state variables in nonlinear filters.
|
||||||
|
|
||||||
|
+ The `conditional_variance_decomposition` option now displays
|
||||||
|
output and stores it as a LaTeX-table when the `TeX` option is
|
||||||
|
invoked,
|
||||||
|
|
||||||
|
+ The `use_calibration` to `estimated_params_init` now also works
|
||||||
|
with ML,
|
||||||
|
|
||||||
|
+ Improved initial estimation checks.
|
||||||
|
|
||||||
|
|
||||||
|
- Steady state
|
||||||
|
|
||||||
|
+ The default solver for finding the steady state is now a
|
||||||
|
trust-region solver (can be triggered explicitly with option
|
||||||
|
`solve_algo=4`),
|
||||||
|
|
||||||
|
+ New options `tolf` and `tolx` to control termination criteria of
|
||||||
|
solver,
|
||||||
|
|
||||||
|
+ The debugging mode now provides the termination values in steady
|
||||||
|
state finding.
|
||||||
|
|
||||||
|
|
||||||
|
- Stochastic simulations
|
||||||
|
|
||||||
|
+ New options `nodecomposition`,
|
||||||
|
|
||||||
|
+ New option `bandpass_filter` to compute bandpass-filtered
|
||||||
|
theoretical and simulated moments,
|
||||||
|
|
||||||
|
+ New option `one_sided_hp_filter` to compute one-sided HP-filtered
|
||||||
|
simulated moments,
|
||||||
|
|
||||||
|
+ `stoch_simul` displays a simulated variance decomposition when
|
||||||
|
simulated moments are requested,
|
||||||
|
|
||||||
|
+ `stoch_simul` saves skewness and kurtosis into respective fields
|
||||||
|
of `oo_` when simulated moments have been requested,
|
||||||
|
|
||||||
|
+ `stoch_simul` saves the unconditional variance decomposition in
|
||||||
|
`oo_.variance_decomposition`,
|
||||||
|
|
||||||
|
+ New option `dr_display_tol` that governs omission of small terms
|
||||||
|
in display of decision rules,
|
||||||
|
|
||||||
|
+ The `stoch_simul` command now prints the displayed tables as LaTeX
|
||||||
|
code when the new `TeX` option is enabled,
|
||||||
|
|
||||||
|
+ The `loglinear` option now works with lagged and leaded exogenous
|
||||||
|
variables like news shocks,
|
||||||
|
|
||||||
|
+ New option `spectral_density` that allows displaying the spectral
|
||||||
|
density of (filtered) endogenous variables,
|
||||||
|
|
||||||
|
+ New option `contemporaneous_correlation` that allows saving
|
||||||
|
contemporaneous correlations in addition to the covariances.
|
||||||
|
|
||||||
|
|
||||||
|
- Identification
|
||||||
|
|
||||||
|
+ New options `diffuse_filter` and `prior_trunc`,
|
||||||
|
|
||||||
|
+ The `identification` command now supports correlations via
|
||||||
|
simulated moments,
|
||||||
|
|
||||||
|
|
||||||
|
- Sensitivity analysis
|
||||||
|
|
||||||
|
+ New blocks `irf_calibration` and `moment_calibration`,
|
||||||
|
|
||||||
|
+ Outputs LaTeX tables if the new `TeX` option is used,
|
||||||
|
|
||||||
|
+ New option `relative_irf` to `irf_calibration` block.
|
||||||
|
|
||||||
|
|
||||||
|
- Conditional forecast
|
||||||
|
|
||||||
|
+ Command `conditional_forecast` now takes into account `histval`
|
||||||
|
block if present.
|
||||||
|
|
||||||
|
|
||||||
|
- Shock decomposition
|
||||||
|
|
||||||
|
+ New option `colormap` to `shocks_decomposition` for controlling
|
||||||
|
the color map used in the shocks decomposition graphs,
|
||||||
|
|
||||||
|
+ `shocks_decomposition` now accepts the `nograph` option,
|
||||||
|
|
||||||
|
+ New command `realtime_shock_decomposition` that for each period `T= [presample,...,nobs]`
|
||||||
|
allows computing the:
|
||||||
|
|
||||||
|
* realtime historical shock decomposition `Y(t|T)`, i.e. without observing data in `[T+1,...,nobs]`
|
||||||
|
|
||||||
|
* forecast shock decomposition `Y(T+k|T)`
|
||||||
|
|
||||||
|
* realtime conditional shock decomposition `Y(T+k|T+k)-Y(T+k|T)`
|
||||||
|
|
||||||
|
+ New block `shock_groups` that allows grouping shocks for the
|
||||||
|
`shock_decomposition` and `realtime_shock_decomposition` commands,
|
||||||
|
|
||||||
|
+ New command `plot_shock_decomposition` that allows plotting the
|
||||||
|
results from `shock_decomposition` and
|
||||||
|
`realtime_shock_decomposition` for different vintages and shock
|
||||||
|
groupings.
|
||||||
|
|
||||||
|
|
||||||
|
- Macroprocessor
|
||||||
|
|
||||||
|
+ Can now pass a macro-variable to the `@#include` macro directive,
|
||||||
|
|
||||||
|
+ New preprocessor flag `-I`, macro directive `@#includepath`, and
|
||||||
|
dynare config file block `[paths]` to pass a search path to the
|
||||||
|
macroprocessor to be used for file inclusion via `@#include`.
|
||||||
|
|
||||||
|
|
||||||
|
- Command line
|
||||||
|
|
||||||
|
+ New option `onlyclearglobals` (do not clear JIT compiled functions
|
||||||
|
with recent versions of Matlab),
|
||||||
|
|
||||||
|
+ New option `minimal_workspace` to use fewer variables in the
|
||||||
|
current workspace,
|
||||||
|
|
||||||
|
+ New option `params_derivs_order` allows limiting the order of the
|
||||||
|
derivatives with respect to the parameters that are calculated by
|
||||||
|
the preprocessor,
|
||||||
|
|
||||||
|
+ New command line option `mingw` to support the MinGW-w64 C/C++
|
||||||
|
Compiler from TDM-GCC for `use_dll`.
|
||||||
|
|
||||||
|
|
||||||
|
- dates/dseries/reporting classes
|
||||||
|
|
||||||
|
+ New methods `abs`, `cumprod` and `chain`,
|
||||||
|
|
||||||
|
+ New option `tableRowIndent` to `addTable`,
|
||||||
|
|
||||||
|
+ Reporting system revamped and made more efficient, dependency on
|
||||||
|
matlab2tikz has been dropped.
|
||||||
|
|
||||||
|
|
||||||
|
- Optimization algorithms
|
||||||
|
|
||||||
|
+ `mode_compute=2` Uses the simulated annealing as described by
|
||||||
|
Corana et al. (1987),
|
||||||
|
|
||||||
|
+ `mode_compute=101` Uses SOLVEOPT as described by Kuntsevich and
|
||||||
|
Kappel (1997),
|
||||||
|
|
||||||
|
+ `mode_compute=102` Uses `simulannealbnd` from Matlab's Global
|
||||||
|
Optimization Toolbox (if available),
|
||||||
|
|
||||||
|
+ New option `silent_optimizer` to shut off output from mode
|
||||||
|
computing/optimization,
|
||||||
|
|
||||||
|
+ New options `verbosity` and `SaveFiles` to control output and
|
||||||
|
saving of files during mode computing/optimization.
|
||||||
|
|
||||||
|
|
||||||
|
- LaTeX output
|
||||||
|
|
||||||
|
+ New command `write_latex_original_model`,
|
||||||
|
|
||||||
|
+ New option `write_equation_tags` to `write_latex_dynamic_model`
|
||||||
|
that allows printing the specified equation tags to the generate
|
||||||
|
LaTeX code,
|
||||||
|
|
||||||
|
+ New command `write_latex_parameter_table` that writes the names and
|
||||||
|
values of model parameters to a LaTeX table,
|
||||||
|
|
||||||
|
+ New command `write_latex_prior_table` that writes the descriptive
|
||||||
|
statistics about the prior distribution to a LaTeX table,
|
||||||
|
|
||||||
|
+ New command `collect_latex_files` that creates one compilable LaTeX
|
||||||
|
file containing all TeX-output.
|
||||||
|
|
||||||
|
|
||||||
|
- Misc.
|
||||||
|
|
||||||
|
+ Provides 64bit preprocessor,
|
||||||
|
|
||||||
|
+ Introduces new path management to avoid conflicts with other
|
||||||
|
toolboxes,
|
||||||
|
|
||||||
|
+ Full compatibility with Matlab 2014b's new graphic interface,
|
||||||
|
|
||||||
|
+ When using `model(linear)`, Dynare automatically checks
|
||||||
|
whether the model is truly linear,
|
||||||
|
|
||||||
|
+ `usedll`, the `msvc` option now supports `normcdf`, `acosh`,
|
||||||
|
`asinh`, and `atanh`,
|
||||||
|
|
||||||
|
+ New parallel option `NumberOfThreadsPerJob` for Windows nodes that
|
||||||
|
sets the number of threads assigned to each remote MATLAB/Octave
|
||||||
|
run,
|
||||||
|
|
||||||
|
+ Improved numerical performance of
|
||||||
|
`schur_statespace_transformation` for very large models,
|
||||||
|
|
||||||
|
+ The `all_values_required` option now also works with `histval`,
|
||||||
|
|
||||||
|
+ Add missing `horizon` option to `ms_forecast`,
|
||||||
|
|
||||||
|
+ BVAR now saves the marginal data density in
|
||||||
|
`oo_.bvar.log_marginal_data_density` and stores prior and
|
||||||
|
posterior information in `oo_.bvar.prior` and
|
||||||
|
`oo_.bvar.posterior`.
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
* Bugs and problems identified in version 4.4.3 and that have been fixed in version 4.5.0:
|
||||||
|
|
||||||
|
|
||||||
|
- BVAR models
|
||||||
|
|
||||||
|
+ `bvar_irf` could display IRFs in an unreadable way when they moved from
|
||||||
|
negative to positive values,
|
||||||
|
|
||||||
|
+ In contrast to what is stated in the documentation, the confidence interval
|
||||||
|
size `conf_sig` was 0.6 by default instead of 0.9.
|
||||||
|
|
||||||
|
|
||||||
|
- Conditional forecasts
|
||||||
|
|
||||||
|
+ The `conditional_forecast` command produced wrong results in calibrated
|
||||||
|
models when used at initial values outside of the steady state (given with
|
||||||
|
`initval`),
|
||||||
|
|
||||||
|
+ The `plot_conditional_forecast` option could produce unreadable figures if
|
||||||
|
the areas overlap,
|
||||||
|
|
||||||
|
+ The `conditional_forecast` command after MLE crashed,
|
||||||
|
|
||||||
|
+ In contrast to what is stated in the manual, the confidence interval size
|
||||||
|
`conf_sig` was 0.6 by default instead of 0.8.
|
||||||
|
|
||||||
|
+ Conditional forecasts were wrong when the declaration of endogenous
|
||||||
|
variables was not preceeding the declaration of the exogenous
|
||||||
|
variables and parameters.
|
||||||
|
|
||||||
|
|
||||||
|
- Discretionary policy
|
||||||
|
|
||||||
|
+ Dynare allowed running models where the number of instruments did not match
|
||||||
|
the number of omitted equations,
|
||||||
|
|
||||||
|
+ Dynare could crash in some cases when trying to display the solution,
|
||||||
|
|
||||||
|
+ Parameter dependence embedded via a `steady_state` was not taken into
|
||||||
|
account, typically resulting in crashes.
|
||||||
|
|
||||||
|
- dseries class
|
||||||
|
|
||||||
|
+ When subtracting a dseries object from a number, the number was instead
|
||||||
|
subtracted from the dseries object.
|
||||||
|
|
||||||
|
|
||||||
|
- DSGE-VAR models
|
||||||
|
|
||||||
|
+ Dynare crashed when estimation encountered non-finite values in the Jacobian
|
||||||
|
at the steady state,
|
||||||
|
|
||||||
|
+ The presence of a constant was not considered for degrees of freedom
|
||||||
|
computation of the Gamma function used during the posterior computation; due
|
||||||
|
to only affecting the constant term, results should be be unaffected, except
|
||||||
|
for model_comparison when comparing models with and without.
|
||||||
|
|
||||||
|
|
||||||
|
- Estimation command
|
||||||
|
|
||||||
|
+ In contrast to what was stated in the manual, the confidence interval size
|
||||||
|
`conf_sig` for `forecast` without MCMC was 0.6 by default instead of 0.9,
|
||||||
|
|
||||||
|
+ Calling estimation after identification could lead to crashes,
|
||||||
|
|
||||||
|
+ When using recursive estimation/forecasting and setting some elements of
|
||||||
|
`nobs` to be larger than the number of observations T in the data,
|
||||||
|
`oo_recursive_` contained additional cell entries that simply repeated the
|
||||||
|
results obtained for `oo_recursive_T`,
|
||||||
|
|
||||||
|
+ Computation of Bayesian smoother could crash for larger models when
|
||||||
|
requesting `forecast` or `filtered_variables`,
|
||||||
|
|
||||||
|
+ Geweke convergence diagnostics were not computed on the full MCMC chain when
|
||||||
|
the `load_mh_file` option was used,
|
||||||
|
|
||||||
|
+ The Geweke convergence diagnostics always used the default `taper_steps` and
|
||||||
|
`geweke_interval`,
|
||||||
|
|
||||||
|
+ Bayesian IRFs (`bayesian_irfs` option) could be displayed in an unreadable
|
||||||
|
way when they move from negative to positive values,
|
||||||
|
|
||||||
|
+ If `bayesian_irfs` was requested when `mh_replic` was too low to compute
|
||||||
|
HPDIs, plotting was crashing,
|
||||||
|
|
||||||
|
+ The x-axis value in `oo_.prior_density` for the standard deviation and
|
||||||
|
correlation of measurement errors was written into a field
|
||||||
|
`mearsurement_errors_*` instead of `measurement_errors_*`,
|
||||||
|
|
||||||
|
+ Using a user-defined `mode_compute` crashed estimation,
|
||||||
|
|
||||||
|
+ Option `mode_compute=10` did not work with infinite prior bounds,
|
||||||
|
|
||||||
|
+ The posterior variances and covariances computed by `moments_varendo` were
|
||||||
|
wrong for very large models due to a matrix erroneously being filled up with
|
||||||
|
zeros,
|
||||||
|
|
||||||
|
+ Using the `forecast` option with `loglinear` erroneously added the unlogged
|
||||||
|
steady state,
|
||||||
|
|
||||||
|
+ When using the `loglinear` option the check for the presence of a constant
|
||||||
|
was erroneously based on the unlogged steady state,
|
||||||
|
|
||||||
|
+ Estimation of `observation_trends` was broken as the trends specified as a
|
||||||
|
function of deep parameters were not correctly updated during estimation,
|
||||||
|
|
||||||
|
+ When using `analytic_derivation`, the parameter values were not set before
|
||||||
|
testing whether the steady state file changes parameter values, leading to
|
||||||
|
subsequent crashes,
|
||||||
|
|
||||||
|
+ If the steady state of an initial parameterization did not solve, the
|
||||||
|
observation equation could erroneously feature no constant when the
|
||||||
|
`use_calibration` option was used,
|
||||||
|
|
||||||
|
+ When computing posterior moments, Dynare falsely displayed that moment
|
||||||
|
computations are skipped, although the computation was performed correctly,
|
||||||
|
|
||||||
|
+ If `conditional_variance_decomposition` was requested, although all
|
||||||
|
variables contain unit roots, Dynare crashed instead of providing an error
|
||||||
|
message,
|
||||||
|
|
||||||
|
+ Computation of the posterior parameter distribution was erroneously based
|
||||||
|
on more draws than specified (there was one additional draw for every Markov
|
||||||
|
chain),
|
||||||
|
|
||||||
|
+ The estimation option `lyapunov=fixed_point` was broken,
|
||||||
|
|
||||||
|
+ Computation of `filtered_vars` with only one requested step crashed Dynare,
|
||||||
|
|
||||||
|
+ Option `kalman_algo=3` was broken with non-diagonal measurement error,
|
||||||
|
|
||||||
|
+ When using the diffuse Kalman filter with missing observations, an additive
|
||||||
|
factor log(2*pi) was missing in the last iteration step,
|
||||||
|
|
||||||
|
+ Passing of the `MaxFunEvals` and `InitialSimplexSize` options to
|
||||||
|
`mode_compute=8` was broken,
|
||||||
|
|
||||||
|
+ Bayesian forecasts contained initial conditions and had the wrong length in
|
||||||
|
both plots and stored variables,
|
||||||
|
|
||||||
|
+ Filtered variables obtained with `mh_replic=0`, ML, or
|
||||||
|
`calibrated_smoother` were padded with zeros at the beginning and end and
|
||||||
|
had the wrong length in stored variables,
|
||||||
|
|
||||||
|
+ Computation of smoothed measurement errors in Bayesian estimation was broken,
|
||||||
|
|
||||||
|
+ The `selected_variables_only` option (`mh_replic=0`, ML, or
|
||||||
|
`calibrated_smoother`) returned wrong results for smoothed, updated, and
|
||||||
|
filtered variables,
|
||||||
|
|
||||||
|
+ Combining the `selected_variables_only` option with forecasts obtained
|
||||||
|
using `mh_replic=0`, ML, or `calibrated_smoother` leaded to crashes,
|
||||||
|
|
||||||
|
+ `oo_.UpdatedVariables` was only filled when the `filtered_vars` option was specified,
|
||||||
|
|
||||||
|
+ When using Bayesian estimation with `filtered_vars`, but without
|
||||||
|
`smoother`, then `oo_.FilteredVariables` erroneously also contained filtered
|
||||||
|
variables at the posterior mean as with `mh_replic=0`,
|
||||||
|
|
||||||
|
+ Running an MCMC a second time in the same folder with a different number of
|
||||||
|
iterations could result in crashes due to the loading of stale files,
|
||||||
|
|
||||||
|
+ Results displayed after Bayesian estimation when not specifying
|
||||||
|
the `smoother` option were based on the parameters at the mode
|
||||||
|
from mode finding instead of the mean parameters from the
|
||||||
|
posterior draws. This affected the smoother results displayed, but
|
||||||
|
also calls to subsequent command relying on the parameters stored
|
||||||
|
in `M_.params` like `stoch_simul`,
|
||||||
|
|
||||||
|
+ The content of `oo_.posterior_std` after Bayesian estimation was based on
|
||||||
|
the standard deviation at the posterior mode, not the one from the MCMC, this
|
||||||
|
was not consistent with the reference manual,
|
||||||
|
|
||||||
|
+ When the initialization of an MCMC run failed, the metropolis.log file was
|
||||||
|
locked, requiring a restart of Matlab to restart estimation,
|
||||||
|
|
||||||
|
+ If the posterior mode was right at the corner of the prior bounds, the
|
||||||
|
initialization of the MCMC erroneously crashed,
|
||||||
|
|
||||||
|
+ If the number of dropped draws via `mh_drop` coincided with the number of
|
||||||
|
draws in a `_mh'-file`, `oo_.posterior.metropolis.mean` and
|
||||||
|
`oo_.posterior.metropolis.Variance` were NaN.
|
||||||
|
|
||||||
|
|
||||||
|
- Estimation and calibrated smoother
|
||||||
|
|
||||||
|
+ When using `observation_trends` with the `prefilter` option, the mean shift
|
||||||
|
due to the trend was not accounted for,
|
||||||
|
|
||||||
|
+ When using `first_obs`>1, the higher trend starting point of
|
||||||
|
`observation_trends` was not taken into account, leading, among other things,
|
||||||
|
to problems in recursive forecasting,
|
||||||
|
|
||||||
|
+ The diffuse Kalman smoother was crashing if the forecast error variance
|
||||||
|
matrix becomes singular,
|
||||||
|
|
||||||
|
+ The multivariate Kalman smoother provided incorrect state estimates when
|
||||||
|
all data for one observation are missing,
|
||||||
|
|
||||||
|
+ The multivariate diffuse Kalman smoother provided incorrect state estimates
|
||||||
|
when the `Finf` matrix becomes singular,
|
||||||
|
|
||||||
|
+ The univariate diffuse Kalman filter was crashing if the initial covariance
|
||||||
|
matrix of the nonstationary state vector is singular,
|
||||||
|
|
||||||
|
|
||||||
|
- Forecats
|
||||||
|
|
||||||
|
+ In contrast to what is stated in the manual, the confidence interval size
|
||||||
|
`conf_sig` was 0.6 by default instead of 0.9.
|
||||||
|
|
||||||
|
+ Forecasting with exogenous deterministic variables provided wrong decision
|
||||||
|
rules, yielding wrong forecasts.
|
||||||
|
|
||||||
|
+ Forecasting with exogenous deterministic variables crashed when the
|
||||||
|
`periods` option was not explicitly specified,
|
||||||
|
|
||||||
|
+ Option `forecast` when used with `initval` was using the initial values in
|
||||||
|
the `initval` block and not the steady state computed from these initial
|
||||||
|
values as the starting point of forecasts.
|
||||||
|
|
||||||
|
|
||||||
|
- Global Sensitivity Analysis
|
||||||
|
|
||||||
|
+ Sensitivity with ML estimation could result in crashes,
|
||||||
|
|
||||||
|
+ Option `mc` must be forced if `neighborhood_width` is used,
|
||||||
|
|
||||||
|
+ Fixed dimension of `stock_logpo` and `stock_ys`,
|
||||||
|
|
||||||
|
+ Incomplete variable initialization could lead to crashes with `prior_range=1`.
|
||||||
|
|
||||||
|
|
||||||
|
- Indentification
|
||||||
|
|
||||||
|
+ Identification did not correctly pass the `lik_init` option,
|
||||||
|
requiring the manual setting of `options_.diffuse_filter=1` in
|
||||||
|
case of unit roots,
|
||||||
|
|
||||||
|
+ Testing identification of standard deviations as the only
|
||||||
|
parameters to be estimated with ML leaded to crashes,
|
||||||
|
|
||||||
|
+ Automatic increase of the lag number for autocovariances when the
|
||||||
|
number of parameters is bigger than the number of non-zero moments
|
||||||
|
was broken,
|
||||||
|
|
||||||
|
+ When using ML, the asymptotic Hessian was not computed,
|
||||||
|
|
||||||
|
+ Checking for singular values when the eigenvectors contained only
|
||||||
|
one column did not work correctly,
|
||||||
|
|
||||||
|
|
||||||
|
- Model comparison
|
||||||
|
|
||||||
|
+ Selection of the `modifiedharmonicmean` estimator was broken,
|
||||||
|
|
||||||
|
|
||||||
|
- Optimal Simple Rules
|
||||||
|
|
||||||
|
+ When covariances were specified, variables that only entered with
|
||||||
|
their variance and no covariance term obtained a wrong weight,
|
||||||
|
resulting in wrong results,
|
||||||
|
|
||||||
|
+ Results reported for stochastic simulations after `osr` were based
|
||||||
|
on the last parameter vector encountered during optimization,
|
||||||
|
which does not necessarily coincide with the optimal parameter
|
||||||
|
vector,
|
||||||
|
|
||||||
|
+ Using only one (co)variance in the objective function resulted in crashes,
|
||||||
|
|
||||||
|
+ For models with non-stationary variables the objective function was computed wrongly.
|
||||||
|
|
||||||
|
|
||||||
|
- Ramsey policy
|
||||||
|
|
||||||
|
+ If a Lagrange multiplier appeared in the model with a lead or a lag
|
||||||
|
of more than one period, the steady state could be wrong.
|
||||||
|
|
||||||
|
+ When using an external steady state file, incorrect steady states
|
||||||
|
could be accepted,
|
||||||
|
|
||||||
|
+ When using an external steady state file with more than one
|
||||||
|
instrument, Dynare crashed,
|
||||||
|
|
||||||
|
+ When using an external steady state file and running `stoch_simul`
|
||||||
|
after `ramsey_planner`, an incorrect steady state was used,
|
||||||
|
|
||||||
|
+ When the number of instruments was not equal to the number of
|
||||||
|
omitted equations, Dynare crashed with a cryptic message,
|
||||||
|
|
||||||
|
+ The `planner_objective` accepted `varexo`, but ignored them for computations,
|
||||||
|
|
||||||
|
|
||||||
|
- Shock decomposition
|
||||||
|
|
||||||
|
+ Did not work with the `parameter_set=calibration` option if an
|
||||||
|
`estimated_params` block is present,
|
||||||
|
|
||||||
|
+ Crashed after MLE.
|
||||||
|
|
||||||
|
|
||||||
|
- Perfect foresight models
|
||||||
|
|
||||||
|
+ The perfect foresight solver could accept a complex solution
|
||||||
|
instead of continuing to look for a real-valued one,
|
||||||
|
|
||||||
|
+ The `initval_file` command only accepted column and not row vectors,
|
||||||
|
|
||||||
|
+ The `initval_file` command did not work with Excel files,
|
||||||
|
|
||||||
|
+ Deterministic simulations with one boundary condition crashed in
|
||||||
|
`solve_one_boundary` due to a missing underscore when passing
|
||||||
|
`options_.simul.maxit`,
|
||||||
|
|
||||||
|
+ Deterministic simulation with exogenous variables lagged by more
|
||||||
|
than one period crashed,
|
||||||
|
|
||||||
|
+ Termination criterion `maxit` was hard-coded for `solve_algo=0`
|
||||||
|
and could no be changed,
|
||||||
|
|
||||||
|
+ When using `block`/`bytecode`, relational operators could not be enforced,
|
||||||
|
|
||||||
|
+ When using `block` some exceptions were not properly handled,
|
||||||
|
leading to code crashes,
|
||||||
|
|
||||||
|
+ Using `periods=1` crashed the solver (bug only partially fixed).
|
||||||
|
|
||||||
|
|
||||||
|
- Smoothing
|
||||||
|
|
||||||
|
+ The univariate Kalman smoother returned wrong results when used
|
||||||
|
with correlated measurement error,
|
||||||
|
|
||||||
|
+ The diffuse smoother sometimes returned linear combinations of the
|
||||||
|
smoothed stochastic trend estimates instead of the original trend
|
||||||
|
estimates.
|
||||||
|
|
||||||
|
- Perturbation reduced form
|
||||||
|
|
||||||
|
+ In contrast to what is stated in the manual, the results of the
|
||||||
|
unconditional variance decomposition were only stored in
|
||||||
|
`oo_.gamma_y(nar+2)`, not in `oo_.variance_decomposition`,
|
||||||
|
|
||||||
|
+ Dynare could crash when the steady state could not be computed
|
||||||
|
when using the `loglinear` option,
|
||||||
|
|
||||||
|
+ Using `bytcode` when declared exogenous variables were not
|
||||||
|
used in the model leaded to crashes in stochastic simulations,
|
||||||
|
|
||||||
|
+ Displaying decision rules involving lags of auxiliary variables of
|
||||||
|
type 0 (leads>1) crashed.
|
||||||
|
|
||||||
|
+ The `relative_irf` option resulted in wrong output at `order>1` as
|
||||||
|
it implicitly relies on linearity.
|
||||||
|
|
||||||
|
|
||||||
|
- Displaying of the MH-history with the `internals` command crashed
|
||||||
|
if parameter names did not have same length.
|
||||||
|
|
||||||
|
- Dynare crashed when the user-defined steady state file returned an
|
||||||
|
error code, but not an conformable-sized steady state vector.
|
||||||
|
|
||||||
|
- Due to a bug in `mjdgges.mex` unstable parameter draws with
|
||||||
|
eigenvalues up to 1+1e-6 could be accepted as stable for the
|
||||||
|
purpose of the Blanchard-Kahn conditions, even if `qz_criterium<1`.
|
||||||
|
|
||||||
|
- The `use_dll` option on Octave for Windows required to pass a
|
||||||
|
compiler flag at the command line, despite the manual stating this
|
||||||
|
was not necessary.
|
||||||
|
|
||||||
|
- Dynare crashed for models with `block` option if the Blanchard-Kahn
|
||||||
|
conditions were not satisfied instead of generating an error
|
||||||
|
message.
|
||||||
|
|
||||||
|
- The `verbose` option did not work with `model(block)`.
|
||||||
|
|
||||||
|
- When falsely specifying the `model(linear)` for nonlinear models,
|
||||||
|
incorrect steady states were accepted instead of aborting.
|
||||||
|
|
||||||
|
- The `STEADY_STATE` operator called on model local variables
|
||||||
|
(so-called pound variables) did not work as expected.
|
||||||
|
|
||||||
|
- The substring operator in macro-processor was broken. The
|
||||||
|
characters of the substring could be mixed with random characters
|
||||||
|
from the memory space.
|
||||||
|
|
||||||
|
- Block decomposition could sometimes cause the preprocessor to crash.
|
||||||
|
|
||||||
|
- A bug when external functions were used in model local variables
|
||||||
|
that were contained in equations that required auxiliary
|
||||||
|
variable/equations led to crashes of Matlab.
|
||||||
|
|
||||||
|
- Sampling from the prior distribution for an inverse gamma II
|
||||||
|
distribution when `prior_trunc>0` could result in incorrect
|
||||||
|
sampling.
|
||||||
|
|
||||||
|
- Sampling from the prior distribution for a uniform distribution
|
||||||
|
when `prior_trunc>0` was ignoring the prior truncation.
|
||||||
|
|
||||||
|
- Conditional forecasts were wrong when the declaration of endogenous
|
||||||
|
variables was not preceeding the declaration of the exogenous
|
||||||
|
variables and parameters.
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
Announcement for Dynare 4.4.3 (on 2014-07-31)
|
Announcement for Dynare 4.4.3 (on 2014-07-31)
|
||||||
=============================================
|
=============================================
|
||||||
|
|
||||||
|
@ -988,7 +1785,7 @@ Here is a list of the main bugfixes since version 4.2.0:
|
||||||
|
|
||||||
* Option `conditional_variance_decomposition' of `stoch_simul' and
|
* Option `conditional_variance_decomposition' of `stoch_simul' and
|
||||||
`estimation' has been fixed
|
`estimation' has been fixed
|
||||||
|
|
||||||
* Automatic detrending now works in conjunction with the `EXPECTATION'
|
* Automatic detrending now works in conjunction with the `EXPECTATION'
|
||||||
operator
|
operator
|
||||||
|
|
||||||
|
@ -1029,7 +1826,7 @@ This release is compatible with MATLAB versions ranging from 6.5 (R13) to 7.11
|
||||||
|
|
||||||
Here is the list of major user-visible changes:
|
Here is the list of major user-visible changes:
|
||||||
|
|
||||||
* New solution algorithms:
|
* New solution algorithms:
|
||||||
|
|
||||||
- Pruning for second order simulations has been added, as described in Kim,
|
- Pruning for second order simulations has been added, as described in Kim,
|
||||||
Kim, Schaumburg and Sims (2008) [1,2]
|
Kim, Schaumburg and Sims (2008) [1,2]
|
||||||
|
@ -1066,7 +1863,7 @@ Here is the list of major user-visible changes:
|
||||||
|
|
||||||
- Syntax of deterministic shocks has changed: after the values keyword,
|
- Syntax of deterministic shocks has changed: after the values keyword,
|
||||||
arbitrary expressions must be enclosed within parentheses (but numeric
|
arbitrary expressions must be enclosed within parentheses (but numeric
|
||||||
constants are still accepted as is)
|
constants are still accepted as is)
|
||||||
|
|
||||||
* Various improvements:
|
* Various improvements:
|
||||||
|
|
||||||
|
@ -1095,7 +1892,7 @@ Here is the list of major user-visible changes:
|
||||||
from the console, it will replace graphical waitbars by text waitbars for
|
from the console, it will replace graphical waitbars by text waitbars for
|
||||||
long computations
|
long computations
|
||||||
|
|
||||||
- Steady option "solve_algo=0" (uses fsolve()) now works under Octave
|
- Steady option "solve_algo=0" (uses fsolve()) now works under Octave
|
||||||
|
|
||||||
* For Emacs users:
|
* For Emacs users:
|
||||||
|
|
||||||
|
|
|
@ -303,7 +303,7 @@ After this, prepare the source and configure the build tree as described for Lin
|
||||||
- **NB**: If not compiling Dynare mex files for Octave, add ```--without-octave``` to the installation command
|
- **NB**: If not compiling Dynare mex files for Octave, add ```--without-octave``` to the installation command
|
||||||
- **NB**: To compile the latest stable version of dynare, follow the same instructions as above, omitting the ```--HEAD``` argument
|
- **NB**: To compile the latest stable version of dynare, follow the same instructions as above, omitting the ```--HEAD``` argument
|
||||||
- **NB**: To update a ```--HEAD``` install of dynare you need to uninstall it then install it again: ```brew uninstall dynare; brew install dynare --HEAD```.
|
- **NB**: To update a ```--HEAD``` install of dynare you need to uninstall it then install it again: ```brew uninstall dynare; brew install dynare --HEAD```.
|
||||||
- **NB**: If you want to maintain a separate git directory of dynare, you can do a ```--HEAD``` install of dynare, then uninstall it. This will have the effect of bringing in all the dependencies you will need to then compile dynare from your git directory. Then, change to the git directory and type:
|
- **NB**: If you want to maintain a separate git directory of dynare, you can do a ```--HEAD``` install of dynare, then uninstall it. This will have the effect of bringing in all the dependencies you will need to then compile dynare from your git directory. (For `flex` and `bison` it may be necessary to symlink them via `brew link bison --force` and `brew link flex --force` as they are keg-only). Then, change to the git directory and type:
|
||||||
- ```autoreconf -si; ./configure --with-matlab=/Applications/MATLAB_R2015a.app MATLAB_VERSION=R2015a```, adjusting the Matlab path and version to accord with your version
|
- ```autoreconf -si; ./configure --with-matlab=/Applications/MATLAB_R2015a.app MATLAB_VERSION=R2015a```, adjusting the Matlab path and version to accord with your version
|
||||||
- Once compilation is done, open Matlab and type the last line shown when you type ```brew info dynare``` in the Terminal window. With the typical Homebrew setup, this is:
|
- Once compilation is done, open Matlab and type the last line shown when you type ```brew info dynare``` in the Terminal window. With the typical Homebrew setup, this is:
|
||||||
- ```addpath /usr/local/opt/dynare/lib/dynare/matlab```
|
- ```addpath /usr/local/opt/dynare/lib/dynare/matlab```
|
||||||
|
|
|
@ -0,0 +1 @@
|
||||||
|
@PACKAGE_VERSION@
|
|
@ -18,7 +18,7 @@ dnl You should have received a copy of the GNU General Public License
|
||||||
dnl along with Dynare. If not, see <http://www.gnu.org/licenses/>.
|
dnl along with Dynare. If not, see <http://www.gnu.org/licenses/>.
|
||||||
|
|
||||||
AC_PREREQ([2.62])
|
AC_PREREQ([2.62])
|
||||||
AC_INIT([dynare], [4.5-unstable])
|
AC_INIT([dynare], [4.6-unstable])
|
||||||
AC_CONFIG_SRCDIR([preprocessor/DynareMain.cc])
|
AC_CONFIG_SRCDIR([preprocessor/DynareMain.cc])
|
||||||
AM_INIT_AUTOMAKE([1.11 -Wall -Wno-portability foreign no-dist-gzip dist-xz tar-pax])
|
AM_INIT_AUTOMAKE([1.11 -Wall -Wno-portability foreign no-dist-gzip dist-xz tar-pax])
|
||||||
|
|
||||||
|
@ -172,6 +172,7 @@ esac
|
||||||
AX_PTHREAD
|
AX_PTHREAD
|
||||||
|
|
||||||
AC_CONFIG_FILES([Makefile
|
AC_CONFIG_FILES([Makefile
|
||||||
|
VERSION
|
||||||
preprocessor/macro/Makefile
|
preprocessor/macro/Makefile
|
||||||
preprocessor/Makefile
|
preprocessor/Makefile
|
||||||
doc/Makefile
|
doc/Makefile
|
||||||
|
|
|
@ -12,7 +12,7 @@
|
||||||
\begin{document}
|
\begin{document}
|
||||||
|
|
||||||
\title{BVAR models ``\`a la Sims'' in Dynare\thanks{Copyright \copyright~2007--2015 S\'ebastien
|
\title{BVAR models ``\`a la Sims'' in Dynare\thanks{Copyright \copyright~2007--2015 S\'ebastien
|
||||||
Villemot; \copyright~2016 S\'ebastien
|
Villemot; \copyright~2016--2017 S\'ebastien
|
||||||
Villemot and Johannes Pfeifer. Permission is granted to copy, distribute and/or modify
|
Villemot and Johannes Pfeifer. Permission is granted to copy, distribute and/or modify
|
||||||
this document under the terms of the GNU Free Documentation
|
this document under the terms of the GNU Free Documentation
|
||||||
License, Version 1.3 or any later version published by the Free
|
License, Version 1.3 or any later version published by the Free
|
||||||
|
@ -27,7 +27,7 @@
|
||||||
|
|
||||||
\author{S\'ebastien Villemot\thanks{Paris School of Economics and
|
\author{S\'ebastien Villemot\thanks{Paris School of Economics and
|
||||||
CEPREMAP.} \and Johannes Pfeifer\thanks{University of Cologne. E-mail: \href{mailto:jpfeifer@uni-koeln.de}{\texttt{jpfeifer@uni-koeln.de}}.}}
|
CEPREMAP.} \and Johannes Pfeifer\thanks{University of Cologne. E-mail: \href{mailto:jpfeifer@uni-koeln.de}{\texttt{jpfeifer@uni-koeln.de}}.}}
|
||||||
\date{First version: September 2007 \hspace{1cm} This version: October 2016}
|
\date{First version: September 2007 \hspace{1cm} This version: May 2017}
|
||||||
|
|
||||||
\maketitle
|
\maketitle
|
||||||
|
|
||||||
|
@ -545,7 +545,7 @@ Most results are stored for future use:
|
||||||
The syntax for computing impulse response functions is:
|
The syntax for computing impulse response functions is:
|
||||||
|
|
||||||
\medskip
|
\medskip
|
||||||
\texttt{bvar\_irf(}\textit{number\_of\_periods},\textit{identification\_scheme}\texttt{);}
|
\texttt{bvar\_irf(}\textit{number\_of\_lags},\textit{identification\_scheme}\texttt{);}
|
||||||
\medskip
|
\medskip
|
||||||
|
|
||||||
The \textit{identification\_scheme} option has two potential values
|
The \textit{identification\_scheme} option has two potential values
|
||||||
|
@ -556,7 +556,25 @@ The \textit{identification\_scheme} option has two potential values
|
||||||
|
|
||||||
Keep in mind that the first factorization of the covariance matrix is sensible to the ordering of the variables (as declared in the mod file with \verb+var+). This is not the case of the second factorization, but its structural interpretation is, at best, unclear (the Matrix square root of a covariance matrix, $\Sigma$, is the unique symmetric matrix $A$ such that $\Sigma = AA$).\newline
|
Keep in mind that the first factorization of the covariance matrix is sensible to the ordering of the variables (as declared in the mod file with \verb+var+). This is not the case of the second factorization, but its structural interpretation is, at best, unclear (the Matrix square root of a covariance matrix, $\Sigma$, is the unique symmetric matrix $A$ such that $\Sigma = AA$).\newline
|
||||||
|
|
||||||
The mean, median, variance and confidence intervals for IRFs are saved in \texttt{oo\_.bvar.irf}
|
If you want to change the length of the IRFs plotted by the command, you can put\\
|
||||||
|
|
||||||
|
\medskip
|
||||||
|
\texttt{options\_.irf=40;}\\
|
||||||
|
\medskip
|
||||||
|
|
||||||
|
before the \texttt{bvar\_irf}-command. Similarly, to change the coverage of the highest posterior density intervals to e.g. 60\% you can put the command\\
|
||||||
|
|
||||||
|
\medskip
|
||||||
|
\texttt{options\_.bvar.conf\_sig=0.6;}\\
|
||||||
|
\medskip
|
||||||
|
|
||||||
|
there.\newline
|
||||||
|
|
||||||
|
|
||||||
|
The mean, median, variance, and confidence intervals for IRFs are saved in \texttt{oo\_.bvar.irf}
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
\section{Examples}
|
\section{Examples}
|
||||||
|
|
||||||
|
|
|
@ -6114,7 +6114,7 @@ singularity is encountered, Dynare by default automatically switches to the univ
|
||||||
recursions as described by @cite{Herbst, 2015}. This setting is only used with
|
recursions as described by @cite{Herbst, 2015}. This setting is only used with
|
||||||
@code{kalman_algo=1} or @code{kalman_algo=3}. In case of using the diffuse Kalman
|
@code{kalman_algo=1} or @code{kalman_algo=3}. In case of using the diffuse Kalman
|
||||||
filter (@code{kalman_algo=3/lik_init=3}), the observables must be stationary. This option
|
filter (@code{kalman_algo=3/lik_init=3}), the observables must be stationary. This option
|
||||||
is not yet compatible with @code{analytical_derivation}.
|
is not yet compatible with @ref{analytic_derivation}.
|
||||||
|
|
||||||
@item kalman_tol = @var{DOUBLE}
|
@item kalman_tol = @var{DOUBLE}
|
||||||
@anchor{kalman_tol} Numerical tolerance for determining the singularity of the covariance matrix of the prediction errors during the Kalman filter (minimum allowed reciprocal of the matrix condition number). Default value is @code{1e-10}
|
@anchor{kalman_tol} Numerical tolerance for determining the singularity of the covariance matrix of the prediction errors during the Kalman filter (minimum allowed reciprocal of the matrix condition number). Default value is @code{1e-10}
|
||||||
|
@ -6260,9 +6260,10 @@ where the model is ill-behaved. By default the original objective function is
|
||||||
used.
|
used.
|
||||||
|
|
||||||
@item analytic_derivation
|
@item analytic_derivation
|
||||||
|
@anchor{analytic_derivation}
|
||||||
Triggers estimation with analytic gradient. The final hessian is also
|
Triggers estimation with analytic gradient. The final hessian is also
|
||||||
computed analytically. Only works for stationary models without
|
computed analytically. Only works for stationary models without
|
||||||
missing observations.
|
missing observations, i.e. for @code{kalman_algo<3}.
|
||||||
|
|
||||||
@item ar = @var{INTEGER}
|
@item ar = @var{INTEGER}
|
||||||
@xref{ar}. Only useful in conjunction with option @code{moments_varendo}.
|
@xref{ar}. Only useful in conjunction with option @code{moments_varendo}.
|
||||||
|
|
|
@ -101,7 +101,11 @@ void SystemResources::getRUS(double& load_avg, long int& pg_avail,
|
||||||
majflt = -1;
|
majflt = -1;
|
||||||
#endif
|
#endif
|
||||||
|
|
||||||
#if !defined(__MINGW32__) && !defined(__CYGWIN32__) && !defined(__CYGWIN__) && !defined(__MINGW64__) && !defined(__CYGWIN64__)
|
|
||||||
|
#define MINGCYGTMP (!defined(__MINGW32__) && !defined(__CYGWIN32__) && !defined(__CYGWIN__))
|
||||||
|
#define MINGCYG (MINGCYGTMP && !defined(__MINGW64__) && !defined(__CYGWIN64__))
|
||||||
|
|
||||||
|
#if MINGCYG
|
||||||
getloadavg(&load_avg, 1);
|
getloadavg(&load_avg, 1);
|
||||||
#else
|
#else
|
||||||
load_avg = -1.0;
|
load_avg = -1.0;
|
||||||
|
|
11
license.txt
11
license.txt
|
@ -1,16 +1,15 @@
|
||||||
Format: http://www.debian.org/doc/packaging-manuals/copyright-format/1.0/
|
Format: http://www.debian.org/doc/packaging-manuals/copyright-format/1.0/
|
||||||
Upstream-Name: Dynare
|
Upstream-Name: Dynare
|
||||||
Upstream-Contact: Dynare Team, whose members in 2016 are:
|
Upstream-Contact: Dynare Team, whose members in 2017 are:
|
||||||
Stéphane Adjemian <stephane.adjemian@univ-lemans.fr>
|
Stéphane Adjemian <stephane.adjemian@univ-lemans.fr>
|
||||||
Houtan Bastani <houtan@dynare.org>
|
Houtan Bastani <houtan@dynare.org>
|
||||||
Michel Juillard <michel.juillard@mjui.fr>
|
Michel Juillard <michel.juillard@mjui.fr>
|
||||||
Frédéric Karamé <frederic.karame@univ-lemans.fr>
|
Frédéric Karamé <frederic.karame@univ-lemans.fr>
|
||||||
Junior Maih <junior.maih@gmail.com>
|
Junior Maih <junior.maih@gmail.com>
|
||||||
Ferhat Mihoubi <fmihoubi@univ-evry.fr>
|
Ferhat Mihoubi <fmihoubi@univ-evry.fr>
|
||||||
George Perendia <george@perendia.orangehome.co.uk>
|
|
||||||
Johannes Pfeifer <jpfeifer@gmx.de>
|
Johannes Pfeifer <jpfeifer@gmx.de>
|
||||||
Marco Ratto <marco.ratto@jrc.ec.europa.eu>
|
Marco Ratto <marco.ratto@ec.europa.eu>
|
||||||
Sébastien Villemot <sebastien@dynare.org>
|
Sébastien Villemot <sebastien.villemot@sciencespo.fr>
|
||||||
Source: http://www.dynare.org
|
Source: http://www.dynare.org
|
||||||
|
|
||||||
Files: *
|
Files: *
|
||||||
|
@ -205,7 +204,7 @@ License: permissive
|
||||||
Unlimited permission is granted to everyone to use, copy, modify or
|
Unlimited permission is granted to everyone to use, copy, modify or
|
||||||
distribute this software.
|
distribute this software.
|
||||||
|
|
||||||
Files: matlab/utilities\graphics\distinguishable_colors.m
|
Files: matlab/utilities/graphics/distinguishable_colors.m
|
||||||
Copyright 2010-2011 by Timothy E. Holy
|
Copyright 2010-2011 by Timothy E. Holy
|
||||||
All rights reserved.
|
All rights reserved.
|
||||||
Redistribution and use in source and binary forms, with or without
|
Redistribution and use in source and binary forms, with or without
|
||||||
|
@ -228,7 +227,7 @@ CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
|
||||||
ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
|
ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
|
||||||
POSSIBILITY OF SUCH DAMAGE.
|
POSSIBILITY OF SUCH DAMAGE.
|
||||||
|
|
||||||
Files: matlab/utilities\graphics\colorspace.m
|
Files: matlab/utilities/graphics/colorspace.m
|
||||||
Pascal Getreuer 2005-2010
|
Pascal Getreuer 2005-2010
|
||||||
All rights reserved.
|
All rights reserved.
|
||||||
Redistribution and use in source and binary forms, with or without
|
Redistribution and use in source and binary forms, with or without
|
||||||
|
|
|
@ -1,20 +0,0 @@
|
||||||
function display(t)
|
|
||||||
|
|
||||||
% Copyright (C) 2013 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/>.
|
|
||||||
|
|
||||||
fprintf('%s = <dynTimeIndex: %s>\n', inputname(1), int2str(t.index));
|
|
|
@ -1,61 +0,0 @@
|
||||||
function t = dynTimeIndex() % --*-- Unitary tests --*--
|
|
||||||
|
|
||||||
% t = dynTimeIndex()
|
|
||||||
%
|
|
||||||
% Constructor for the dynTimeIndex class.
|
|
||||||
%
|
|
||||||
% INPUTS:
|
|
||||||
% None.
|
|
||||||
%
|
|
||||||
% OUTPUTS:
|
|
||||||
% * t, dynTimeIndex object.
|
|
||||||
%
|
|
||||||
% DESCRIPTION:
|
|
||||||
% The dynTimeIndex object is used to shift backward or forward dseries objects. For instance, if ts
|
|
||||||
% is a dseries object and t is a dynTimeIndex object then the following expressions are equivalent:
|
|
||||||
%
|
|
||||||
% us = ts.lag()
|
|
||||||
% us = ts.lag(1)
|
|
||||||
% us = lag(ts,1)
|
|
||||||
% us = ts(t-1)
|
|
||||||
%
|
|
||||||
% This class has only one member: t = int8(0) when instantiated.
|
|
||||||
|
|
||||||
% Copyright (C) 2013 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/>.
|
|
||||||
|
|
||||||
t = struct();
|
|
||||||
|
|
||||||
t.index = int8(0);
|
|
||||||
|
|
||||||
t = class(t,'dynTimeIndex');
|
|
||||||
|
|
||||||
%@test:1
|
|
||||||
%$ % Instantiate a dynTimeIndex object
|
|
||||||
%$ try
|
|
||||||
%$ u = dynTimeIndex();
|
|
||||||
%$ t(1) = 1;
|
|
||||||
%$ catch
|
|
||||||
%$ t(1) = 0;
|
|
||||||
%$ end
|
|
||||||
%$
|
|
||||||
%$ if t(1)
|
|
||||||
%$ t(2) = isa(u,'dynTimeIndex');
|
|
||||||
%$ end
|
|
||||||
%$
|
|
||||||
%$ T = all(t);
|
|
||||||
%@eof:1
|
|
|
@ -1,62 +0,0 @@
|
||||||
function C = minus(A,B) % --*-- Unitary tests --*--
|
|
||||||
|
|
||||||
% C = minus(A,B)
|
|
||||||
%
|
|
||||||
% Overloads binary minus operator.
|
|
||||||
%
|
|
||||||
% INPUTS:
|
|
||||||
% * A, dynTimeIndex object.
|
|
||||||
% * B, integer scalar.
|
|
||||||
%
|
|
||||||
% OUTPUTS:
|
|
||||||
% * C, dynTimeIndex object.
|
|
||||||
%
|
|
||||||
% EXAMPLE:
|
|
||||||
%
|
|
||||||
% >> t = dynTimeIndex();
|
|
||||||
% >> t.index
|
|
||||||
%
|
|
||||||
% ans =
|
|
||||||
%
|
|
||||||
% 0
|
|
||||||
%
|
|
||||||
% >> s = t-1;
|
|
||||||
% >> s.index
|
|
||||||
%
|
|
||||||
% ans =
|
|
||||||
%
|
|
||||||
% -1
|
|
||||||
%
|
|
||||||
|
|
||||||
% Copyright (C) 2013 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/>.
|
|
||||||
|
|
||||||
if ~(isa(A,'dynTimeIndex') || isint(B))
|
|
||||||
error(['dynTimeIndex::plus: Input arguments ''' inputname(1) ''' and ''' inputname(2) ''' must be a dynTimeIndex object and an integer!'])
|
|
||||||
end
|
|
||||||
|
|
||||||
C = struct();
|
|
||||||
C.index = A.index-B;
|
|
||||||
C = class(C,'dynTimeIndex');
|
|
||||||
|
|
||||||
%@test:1
|
|
||||||
%$ a = dynTimeIndex();
|
|
||||||
%$ b = a-1;
|
|
||||||
%$ t(1) = isa(b,'dynTimeIndex');
|
|
||||||
%$ t(2) = isequal(b.index,-int8(1));
|
|
||||||
%$ T = all(t);
|
|
||||||
%@eof:1
|
|
|
@ -1,74 +0,0 @@
|
||||||
function C = mpower(A,B) % --*-- Unitary tests --*--
|
|
||||||
|
|
||||||
% C = mpower(A,B)
|
|
||||||
%
|
|
||||||
% Overloads binary mpower operator (^).
|
|
||||||
%
|
|
||||||
% INPUTS :
|
|
||||||
% * A, dynTimeIndex object.
|
|
||||||
% * B, integer scalar.
|
|
||||||
%
|
|
||||||
% OUTPUTS :
|
|
||||||
% * C, dynTimeIndex object.
|
|
||||||
%
|
|
||||||
% EXAMPLE 1 :
|
|
||||||
%
|
|
||||||
% >> B = dynTimeIndex()-1;
|
|
||||||
% >> B
|
|
||||||
% B = <dynTimeIndex: -1>
|
|
||||||
% >> B^4
|
|
||||||
% ans = <dynTimeIndex: -4>
|
|
||||||
% >>
|
|
||||||
%
|
|
||||||
% EXAMPLE 2 :
|
|
||||||
% This method can be used to apply the lead and lag methods an arbitrary number of times to a dseries object. For instance, if
|
|
||||||
% ts is a dseries object, and if we define
|
|
||||||
%
|
|
||||||
% >> B = dynTimeIndex()-1;
|
|
||||||
% >> F = dynTimeIndex()+1;
|
|
||||||
%
|
|
||||||
% B and F can be used as lag and lead operators and the following syntax:
|
|
||||||
%
|
|
||||||
% >> us = ts(F^2);
|
|
||||||
%
|
|
||||||
% is equivalent to
|
|
||||||
%
|
|
||||||
% >> us = ts.lead(2)
|
|
||||||
%
|
|
||||||
% or
|
|
||||||
%
|
|
||||||
% >> us = ts.lead.lead
|
|
||||||
%
|
|
||||||
|
|
||||||
% Copyright (C) 2013 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/>.
|
|
||||||
|
|
||||||
if ~(isa(A,'dynTimeIndex') || isint(B))
|
|
||||||
error(['dynTimeIndex::mpower: Input arguments ''' inputname(1) ''' and ''' inputname(2) ''' must be a dynTimeIndex object and an integer!'])
|
|
||||||
end
|
|
||||||
|
|
||||||
C = struct();
|
|
||||||
C.index = A.index*B;
|
|
||||||
C = class(C,'dynTimeIndex');
|
|
||||||
|
|
||||||
%@test:1
|
|
||||||
%$ a = dynTimeIndex()+1;
|
|
||||||
%$ b = a^2;
|
|
||||||
%$ t(1) = isa(b,'dynTimeIndex');
|
|
||||||
%$ t(2) = isequal(b.index,int8(2));
|
|
||||||
%$ T = all(t);
|
|
||||||
%@eof:1
|
|
|
@ -1,62 +0,0 @@
|
||||||
function C = plus(A,B) % --*-- Unitary tests --*--
|
|
||||||
|
|
||||||
% C = plus(A,B)
|
|
||||||
%
|
|
||||||
% Overloads binary plus operator.
|
|
||||||
%
|
|
||||||
% INPUTS:
|
|
||||||
% * A, dynTimeIndex object.
|
|
||||||
% * B, integer scalar.
|
|
||||||
%
|
|
||||||
% OUTPUTS:
|
|
||||||
% * C, dynTimeIndex object.
|
|
||||||
%
|
|
||||||
% EXAMPLE:
|
|
||||||
%
|
|
||||||
% >> t = dynTimeIndex();
|
|
||||||
% >> t.index
|
|
||||||
%
|
|
||||||
% ans =
|
|
||||||
%
|
|
||||||
% 0
|
|
||||||
%
|
|
||||||
% >> s = t+1;
|
|
||||||
% >> s.index
|
|
||||||
%
|
|
||||||
% ans =
|
|
||||||
%
|
|
||||||
% 1
|
|
||||||
%
|
|
||||||
|
|
||||||
% Copyright (C) 2013 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/>.
|
|
||||||
|
|
||||||
if ~(isa(A,'dynTimeIndex') || isint(B))
|
|
||||||
error(['dynTimeIndex::plus: Input arguments ''' inputname(1) ''' and ''' inputname(2) ''' must be a dynTimeIndex object and an integer!'])
|
|
||||||
end
|
|
||||||
|
|
||||||
C = struct();
|
|
||||||
C.index = A.index+B;
|
|
||||||
C = class(C,'dynTimeIndex');
|
|
||||||
|
|
||||||
%@test:1
|
|
||||||
%$ a = dynTimeIndex();
|
|
||||||
%$ b = a+1;
|
|
||||||
%$ t(1) = isa(b,'dynTimeIndex');
|
|
||||||
%$ t(2) = isequal(b.index,int8(1));
|
|
||||||
%$ T = all(t);
|
|
||||||
%@eof:1
|
|
|
@ -1,20 +0,0 @@
|
||||||
function val = subsasgn(val, idx, rhs)
|
|
||||||
|
|
||||||
% Copyright (C) 2013 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/>.
|
|
||||||
|
|
||||||
error('dynTimeIndex::subsasgn: Members of dynTimeIndex class are private!')
|
|
|
@ -1,54 +0,0 @@
|
||||||
function B = subsref(A,S) % --*-- Unitary tests --*--
|
|
||||||
|
|
||||||
% B = subsref(A,S)
|
|
||||||
%
|
|
||||||
% Overloads the subsref method for dynTimeIndex class. This method only allows to get
|
|
||||||
% the value of the field `index`.
|
|
||||||
|
|
||||||
% Copyright (C) 2013 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/>.
|
|
||||||
|
|
||||||
if length(S)>1
|
|
||||||
error('dynTimeIndex::subsref: Something is wrong in your syntax!')
|
|
||||||
end
|
|
||||||
|
|
||||||
if isequal(S.type,'.')
|
|
||||||
if isequal(S.subs,'index')
|
|
||||||
B = builtin('subsref', A, S(1));
|
|
||||||
else
|
|
||||||
error(['dynTimeIndex::subsref: ' S.subs ' is not a known member!'])
|
|
||||||
end
|
|
||||||
else
|
|
||||||
error('dynTimeIndex::subsref: Something is wrong in your syntax!')
|
|
||||||
end
|
|
||||||
|
|
||||||
%@test:1
|
|
||||||
%$ % Instantiate a dynTimeIndex object
|
|
||||||
%$ u = dynTimeIndex();
|
|
||||||
%$ try
|
|
||||||
%$ v = u.index;
|
|
||||||
%$ t(1) = 1;
|
|
||||||
%$ catch
|
|
||||||
%$ t(1) = 0;
|
|
||||||
%$ end
|
|
||||||
%$
|
|
||||||
%$ if t(1)
|
|
||||||
%$ t(2) = isequal(v,int8(0));
|
|
||||||
%$ end
|
|
||||||
%$
|
|
||||||
%$ T = all(t);
|
|
||||||
%@eof:1
|
|
|
@ -8,7 +8,7 @@ function [AHess, DLIK, LIK] = AHessian(T,R,Q,H,P,Y,DT,DYss,DOm,DH,DP,start,mf,ka
|
||||||
% NOTE: the derivative matrices (DT,DR ...) are 3-dim. arrays with last
|
% NOTE: the derivative matrices (DT,DR ...) are 3-dim. arrays with last
|
||||||
% dimension equal to the number of structural parameters
|
% dimension equal to the number of structural parameters
|
||||||
|
|
||||||
% Copyright (C) 2011-2016 Dynare Team
|
% Copyright (C) 2011-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -39,7 +39,7 @@ F_singular = 1;
|
||||||
|
|
||||||
lik = zeros(smpl,1); % Initialization of the vector gathering the densities.
|
lik = zeros(smpl,1); % Initialization of the vector gathering the densities.
|
||||||
LIK = Inf; % Default value of the log likelihood.
|
LIK = Inf; % Default value of the log likelihood.
|
||||||
if nargout > 1,
|
if nargout > 1
|
||||||
DLIK = zeros(k,1); % Initialization of the score.
|
DLIK = zeros(k,1); % Initialization of the score.
|
||||||
end
|
end
|
||||||
AHess = zeros(k,k); % Initialization of the Hessian
|
AHess = zeros(k,k); % Initialization of the Hessian
|
||||||
|
@ -47,7 +47,7 @@ Da = zeros(mm,k); % State vector.
|
||||||
Dv = zeros(length(mf),k);
|
Dv = zeros(length(mf),k);
|
||||||
|
|
||||||
% for ii = 1:k
|
% for ii = 1:k
|
||||||
% DOm = DR(:,:,ii)*Q*transpose(R) + R*DQ(:,:,ii)*transpose(R) + R*Q*transpose(DR(:,:,ii));
|
% DOm = DR(:,:,ii)*Q*transpose(R) + R*DQ(:,:,ii)*transpose(R) + R*Q*transpose(DR(:,:,ii));
|
||||||
% end
|
% end
|
||||||
|
|
||||||
while notsteady && t<smpl
|
while notsteady && t<smpl
|
||||||
|
@ -66,9 +66,7 @@ while notsteady && t<smpl
|
||||||
iF = inv(F);
|
iF = inv(F);
|
||||||
K = P(:,mf)*iF;
|
K = P(:,mf)*iF;
|
||||||
lik(t) = log(det(F))+transpose(v)*iF*v;
|
lik(t) = log(det(F))+transpose(v)*iF*v;
|
||||||
|
|
||||||
[DK,DF,DP1] = computeDKalman(T,DT,DOm,P,DP,DH,mf,iF,K);
|
[DK,DF,DP1] = computeDKalman(T,DT,DOm,P,DP,DH,mf,iF,K);
|
||||||
|
|
||||||
for ii = 1:k
|
for ii = 1:k
|
||||||
Dv(:,ii) = -Da(mf,ii) - DYss(mf,ii);
|
Dv(:,ii) = -Da(mf,ii) - DYss(mf,ii);
|
||||||
Da(:,ii) = DT(:,:,ii)*(a+K*v) + T*(Da(:,ii)+DK(:,:,ii)*v + K*Dv(:,ii));
|
Da(:,ii) = DT(:,:,ii)*(a+K*v) + T*(Da(:,ii)+DK(:,:,ii)*v + K*Dv(:,ii));
|
||||||
|
@ -81,7 +79,7 @@ while notsteady && t<smpl
|
||||||
if t>=start
|
if t>=start
|
||||||
AHess = AHess + Dv'*iF*Dv + .5*(vecDPmf' * kron(iF,iF) * vecDPmf);
|
AHess = AHess + Dv'*iF*Dv + .5*(vecDPmf' * kron(iF,iF) * vecDPmf);
|
||||||
end
|
end
|
||||||
a = T*(a+K*v);
|
a = T*(a+K*v);
|
||||||
P = T*(P-K*P(mf,:))*transpose(T)+Om;
|
P = T*(P-K*P(mf,:))*transpose(T)+Om;
|
||||||
DP = DP1;
|
DP = DP1;
|
||||||
end
|
end
|
||||||
|
@ -107,8 +105,8 @@ if t < smpl
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
if t>=start
|
if t>=start
|
||||||
AHess = AHess + Dv'*iF*Dv;
|
AHess = AHess + Dv'*iF*Dv;
|
||||||
end
|
end
|
||||||
a = T*(a+K*v);
|
a = T*(a+K*v);
|
||||||
lik(t) = transpose(v)*iF*v;
|
lik(t) = transpose(v)*iF*v;
|
||||||
end
|
end
|
||||||
|
@ -124,17 +122,17 @@ if t < smpl
|
||||||
% H(ii,jj) = trace(iPmf*(.5*DP(mf,mf,ii)*iPmf*DP(mf,mf,jj) + Dv(:,ii)*Dv(:,jj)'));
|
% H(ii,jj) = trace(iPmf*(.5*DP(mf,mf,ii)*iPmf*DP(mf,mf,jj) + Dv(:,ii)*Dv(:,jj)'));
|
||||||
% end
|
% end
|
||||||
% end
|
% end
|
||||||
end
|
end
|
||||||
|
|
||||||
AHess = -AHess;
|
AHess = -AHess;
|
||||||
if nargout > 1,
|
if nargout > 1
|
||||||
DLIK = DLIK/2;
|
DLIK = DLIK/2;
|
||||||
end
|
end
|
||||||
% adding log-likelihhod constants
|
% adding log-likelihhod constants
|
||||||
lik = (lik + pp*log(2*pi))/2;
|
lik = (lik + pp*log(2*pi))/2;
|
||||||
|
|
||||||
LIK = sum(lik(start:end)); % Minus the log-likelihood.
|
LIK = sum(lik(start:end)); % Minus the log-likelihood.
|
||||||
% end of main function
|
% end of main function
|
||||||
|
|
||||||
function [DK,DF,DP1] = computeDKalman(T,DT,DOm,P,DP,DH,mf,iF,K)
|
function [DK,DF,DP1] = computeDKalman(T,DT,DOm,P,DP,DH,mf,iF,K)
|
||||||
|
|
||||||
|
@ -142,13 +140,11 @@ k = size(DT,3);
|
||||||
tmp = P-K*P(mf,:);
|
tmp = P-K*P(mf,:);
|
||||||
|
|
||||||
for ii = 1:k
|
for ii = 1:k
|
||||||
DF(:,:,ii) = DP(mf,mf,ii) + DH(:,:,ii);
|
DF(:,:,ii) = DP(mf,mf,ii) + DH(:,:,ii);
|
||||||
DiF(:,:,ii) = -iF*DF(:,:,ii)*iF;
|
DiF(:,:,ii) = -iF*DF(:,:,ii)*iF;
|
||||||
DK(:,:,ii) = DP(:,mf,ii)*iF + P(:,mf)*DiF(:,:,ii);
|
DK(:,:,ii) = DP(:,mf,ii)*iF + P(:,mf)*DiF(:,:,ii);
|
||||||
Dtmp = DP(:,:,ii) - DK(:,:,ii)*P(mf,:) - K*DP(mf,:,ii);
|
Dtmp = DP(:,:,ii) - DK(:,:,ii)*P(mf,:) - K*DP(mf,:,ii);
|
||||||
DP1(:,:,ii) = DT(:,:,ii)*tmp*T' + T*Dtmp*T' + T*tmp*DT(:,:,ii)' + DOm(:,:,ii);
|
DP1(:,:,ii) = DT(:,:,ii)*tmp*T' + T*Dtmp*T' + T*tmp*DT(:,:,ii)' + DOm(:,:,ii);
|
||||||
end
|
end
|
||||||
|
|
||||||
% end of computeDKalman
|
% end of computeDKalman
|
||||||
|
|
||||||
|
|
|
@ -1,4 +1,4 @@
|
||||||
function e = SPAimerr(c);
|
function e = SPAimerr(c)
|
||||||
% e = aimerr(c);
|
% e = aimerr(c);
|
||||||
%
|
%
|
||||||
% Interpret the return codes generated by the aim routines.
|
% Interpret the return codes generated by the aim routines.
|
||||||
|
|
|
@ -8,9 +8,9 @@ function [b,rts,ia,nexact,nnumeric,lgroots,aimcode] = ...
|
||||||
% roots. This procedure will fail if the companion matrix is
|
% roots. This procedure will fail if the companion matrix is
|
||||||
% defective and does not have a linearly independent set of
|
% defective and does not have a linearly independent set of
|
||||||
% eigenvectors associated with the big roots.
|
% eigenvectors associated with the big roots.
|
||||||
%
|
%
|
||||||
% Input arguments:
|
% Input arguments:
|
||||||
%
|
%
|
||||||
% h Structural coefficient matrix (neq,neq*(nlag+1+nlead)).
|
% h Structural coefficient matrix (neq,neq*(nlag+1+nlead)).
|
||||||
% neq Number of equations.
|
% neq Number of equations.
|
||||||
% nlag Number of lags.
|
% nlag Number of lags.
|
||||||
|
@ -19,9 +19,9 @@ function [b,rts,ia,nexact,nnumeric,lgroots,aimcode] = ...
|
||||||
% by numeric_shift and reduced_form.
|
% by numeric_shift and reduced_form.
|
||||||
% uprbnd Inclusive upper bound for the modulus of roots
|
% uprbnd Inclusive upper bound for the modulus of roots
|
||||||
% allowed in the reduced form.
|
% allowed in the reduced form.
|
||||||
%
|
%
|
||||||
% Output arguments:
|
% Output arguments:
|
||||||
%
|
%
|
||||||
% b Reduced form coefficient matrix (neq,neq*nlag).
|
% b Reduced form coefficient matrix (neq,neq*nlag).
|
||||||
% rts Roots returned by eig.
|
% rts Roots returned by eig.
|
||||||
% ia Dimension of companion matrix (number of non-trivial
|
% ia Dimension of companion matrix (number of non-trivial
|
||||||
|
@ -57,7 +57,7 @@ function [b,rts,ia,nexact,nnumeric,lgroots,aimcode] = ...
|
||||||
% pages 472-489
|
% pages 472-489
|
||||||
|
|
||||||
b=[];rts=[];ia=[];nexact=[];nnumeric=[];lgroots=[];aimcode=[];
|
b=[];rts=[];ia=[];nexact=[];nnumeric=[];lgroots=[];aimcode=[];
|
||||||
if(nlag<1 || nlead<1)
|
if(nlag<1 || nlead<1)
|
||||||
error('Aim_eig: model must have at least one lag and one lead');
|
error('Aim_eig: model must have at least one lag and one lead');
|
||||||
end
|
end
|
||||||
% Initialization.
|
% Initialization.
|
||||||
|
@ -66,26 +66,26 @@ bcols=neq*nlag;q=zeros(qrows,qcols);rts=zeros(qcols,1);
|
||||||
[h,q,iq,nexact]=SPExact_shift(h,q,iq,qrows,qcols,neq);
|
[h,q,iq,nexact]=SPExact_shift(h,q,iq,qrows,qcols,neq);
|
||||||
if (iq>qrows)
|
if (iq>qrows)
|
||||||
aimcode = 61;
|
aimcode = 61;
|
||||||
return;
|
return
|
||||||
end
|
end
|
||||||
[h,q,iq,nnumeric]=SPNumeric_shift(h,q,iq,qrows,qcols,neq,condn);
|
[h,q,iq,nnumeric]=SPNumeric_shift(h,q,iq,qrows,qcols,neq,condn);
|
||||||
if (iq>qrows)
|
if (iq>qrows)
|
||||||
aimcode = 62;
|
aimcode = 62;
|
||||||
return;
|
return
|
||||||
end
|
end
|
||||||
[a,ia,js] = SPBuild_a(h,qcols,neq);
|
[a,ia,js] = SPBuild_a(h,qcols,neq);
|
||||||
if (ia ~= 0)
|
if (ia ~= 0)
|
||||||
if any(any(isnan(a))) || any(any(isinf(a)))
|
if any(any(isnan(a))) || any(any(isinf(a)))
|
||||||
disp('A is NAN or INF')
|
disp('A is NAN or INF')
|
||||||
aimcode=63;
|
aimcode=63;
|
||||||
return
|
return
|
||||||
end
|
end
|
||||||
[w,rts,lgroots,flag_trouble]=SPEigensystem(a,uprbnd,min(length(js),qrows-iq+1));
|
[w,rts,lgroots,flag_trouble]=SPEigensystem(a,uprbnd,min(length(js),qrows-iq+1));
|
||||||
if flag_trouble==1;
|
if flag_trouble==1
|
||||||
disp('Problem in SPEIG');
|
disp('Problem in SPEIG');
|
||||||
aimcode=64;
|
aimcode=64;
|
||||||
return
|
return
|
||||||
end
|
end
|
||||||
q = SPCopy_w(q,w,js,iq,qrows);
|
q = SPCopy_w(q,w,js,iq,qrows);
|
||||||
end
|
end
|
||||||
test=nexact+nnumeric+lgroots;
|
test=nexact+nnumeric+lgroots;
|
||||||
|
|
|
@ -51,7 +51,7 @@ a(hrows,:) = hs(:,left);
|
||||||
% Delete inessential lags and build index array js. js indexes the
|
% Delete inessential lags and build index array js. js indexes the
|
||||||
% columns in the big transition matrix that correspond to the
|
% columns in the big transition matrix that correspond to the
|
||||||
% essential lags in the model. They are the columns of q that will
|
% essential lags in the model. They are the columns of q that will
|
||||||
% get the unstable left eigenvectors.
|
% get the unstable left eigenvectors.
|
||||||
|
|
||||||
js = 1:qcols;
|
js = 1:qcols;
|
||||||
zerocols = sum(abs(a)) == 0;
|
zerocols = sum(abs(a)) == 0;
|
||||||
|
|
|
@ -2,7 +2,7 @@ function q = SPCopy_w(q,w,js,iq,qrows)
|
||||||
% q = SPCopy_w(q,w,js,iq,qrows)
|
% q = SPCopy_w(q,w,js,iq,qrows)
|
||||||
%
|
%
|
||||||
% Copy the eigenvectors corresponding to the largest roots into the
|
% Copy the eigenvectors corresponding to the largest roots into the
|
||||||
% remaining empty rows and columns js of q
|
% remaining empty rows and columns js of q
|
||||||
|
|
||||||
% Author: Gary Anderson
|
% Author: Gary Anderson
|
||||||
% Original file downloaded from:
|
% Original file downloaded from:
|
||||||
|
|
|
@ -30,7 +30,7 @@ function [w,rts,lgroots,flag_trouble] = SPEigensystem(a,uprbnd,rowsLeft)
|
||||||
% Journal of Economic Dynamics and Control, 2010, vol. 34, issue 3,
|
% Journal of Economic Dynamics and Control, 2010, vol. 34, issue 3,
|
||||||
% pages 472-489
|
% pages 472-489
|
||||||
|
|
||||||
opts.disp=0;
|
opts.disp=0;
|
||||||
% next block is commented out because eigs() intermitently returns different rts
|
% next block is commented out because eigs() intermitently returns different rts
|
||||||
%try
|
%try
|
||||||
% [w,d] = eigs(a',rowsLeft,'LM',opts);
|
% [w,d] = eigs(a',rowsLeft,'LM',opts);
|
||||||
|
@ -56,7 +56,7 @@ mag = abs(rts);
|
||||||
[mag,k] = sort(-mag);
|
[mag,k] = sort(-mag);
|
||||||
rts = rts(k);
|
rts = rts(k);
|
||||||
%end
|
%end
|
||||||
flag_trouble=0;
|
flag_trouble=0;
|
||||||
|
|
||||||
%ws=SPSparse(w);
|
%ws=SPSparse(w);
|
||||||
ws=sparse(w);
|
ws=sparse(w);
|
||||||
|
@ -65,7 +65,7 @@ ws = ws(:,k);
|
||||||
% Given a complex conjugate pair of vectors W = [w1,w2], there is a
|
% Given a complex conjugate pair of vectors W = [w1,w2], there is a
|
||||||
% nonsingular matrix D such that W*D = real(W) + imag(W). That is to
|
% nonsingular matrix D such that W*D = real(W) + imag(W). That is to
|
||||||
% say, W and real(W)+imag(W) span the same subspace, which is all
|
% say, W and real(W)+imag(W) span the same subspace, which is all
|
||||||
% that aim cares about.
|
% that aim cares about.
|
||||||
|
|
||||||
ws = real(ws) + imag(ws);
|
ws = real(ws) + imag(ws);
|
||||||
|
|
||||||
|
|
|
@ -44,4 +44,3 @@ while( any(zerorows) && iq <= qrows )
|
||||||
zerorows = find( sum(abs( hs(:,right)' ))==0 );
|
zerorows = find( sum(abs( hs(:,right)' ))==0 );
|
||||||
end
|
end
|
||||||
h=full(hs);
|
h=full(hs);
|
||||||
|
|
||||||
|
|
|
@ -2,7 +2,7 @@ function scof = SPObstruct(cof,cofb,neq,nlag,nlead)
|
||||||
% scof = SPObstruct(cof,cofb,neq,nlag,nlead)
|
% scof = SPObstruct(cof,cofb,neq,nlag,nlead)
|
||||||
%
|
%
|
||||||
% Construct the coefficients in the observable structure.
|
% Construct the coefficients in the observable structure.
|
||||||
%
|
%
|
||||||
% Input arguments:
|
% Input arguments:
|
||||||
%
|
%
|
||||||
% cof structural coefficients
|
% cof structural coefficients
|
||||||
|
@ -51,7 +51,7 @@ qs=sparse(q);
|
||||||
qs(1:rc,1:cc) = sparse(cofb);
|
qs(1:rc,1:cc) = sparse(cofb);
|
||||||
qcols = neq*(nlag+nlead);
|
qcols = neq*(nlag+nlead);
|
||||||
|
|
||||||
if( nlead > 1 )
|
if( nlead > 1 )
|
||||||
for i = 1:nlead-1
|
for i = 1:nlead-1
|
||||||
rows = i*neq + (1:neq);
|
rows = i*neq + (1:neq);
|
||||||
qs(rows,:) = SPShiftright( qs((rows-neq),:), neq );
|
qs(rows,:) = SPShiftright( qs((rows-neq),:), neq );
|
||||||
|
|
|
@ -1,4 +1,4 @@
|
||||||
function [nonsing,b] = SPReduced_form(q,qrows,qcols,bcols,neq,condn);
|
function [nonsing,b] = SPReduced_form(q,qrows,qcols,bcols,neq,condn)
|
||||||
% [nonsing,b] = SPReduced_form(q,qrows,qcols,bcols,neq,b,condn);
|
% [nonsing,b] = SPReduced_form(q,qrows,qcols,bcols,neq,b,condn);
|
||||||
%
|
%
|
||||||
% Compute reduced-form coefficient matrix, b.
|
% Compute reduced-form coefficient matrix, b.
|
||||||
|
@ -48,4 +48,3 @@ else %rescale by dividing row by maximal qr element
|
||||||
b = full(b);
|
b = full(b);
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
|
@ -3,7 +3,7 @@ function [y] = SPShiftright(x,n)
|
||||||
% [y] = shiftright(x,n)
|
% [y] = shiftright(x,n)
|
||||||
%
|
%
|
||||||
% Shift the rows of x to the right by n columns, leaving zeros in the
|
% Shift the rows of x to the right by n columns, leaving zeros in the
|
||||||
% first n columns.
|
% first n columns.
|
||||||
|
|
||||||
% Original author: Gary Anderson
|
% Original author: Gary Anderson
|
||||||
% Original file downloaded from:
|
% Original file downloaded from:
|
||||||
|
|
|
@ -1,20 +1,20 @@
|
||||||
function [dr,aimcode,rts]=dynAIMsolver1(jacobia_,M_,dr)
|
function [dr,aimcode,rts]=dynAIMsolver1(jacobia_,M_,dr)
|
||||||
% function [dr,aimcode]=dynAIMsolver1(jacobia_,M_,dr)
|
% function [dr,aimcode]=dynAIMsolver1(jacobia_,M_,dr)
|
||||||
% Maps Dynare jacobia to AIM 1st order model solver designed and developed by Gary ANderson
|
% Maps Dynare jacobia to AIM 1st order model solver designed and developed by Gary ANderson
|
||||||
% and derives the solution for gy=dr.hgx and gu=dr.hgu from the AIM outputs
|
% and derives the solution for gy=dr.hgx and gu=dr.hgu from the AIM outputs
|
||||||
% AIM System is given as a sum:
|
% AIM System is given as a sum:
|
||||||
% i.e. for i=-$...+& SUM(Hi*xt+i)= £*zt, t = 0, . . . ,?
|
% i.e. for i=-$...+& SUM(Hi*xt+i)= £*zt, t = 0, . . . ,?
|
||||||
% and its input as single array of matrices: [H-$... Hi ... H+&]
|
% and its input as single array of matrices: [H-$... Hi ... H+&]
|
||||||
% and its solution as xt=SUM( Bi*xt+i) + @*£*zt for i=-$...-1
|
% and its solution as xt=SUM( Bi*xt+i) + @*£*zt for i=-$...-1
|
||||||
% with the output in form bb=[B-$... Bi ... B-1] and @=inv(Ho+H1*B-1)
|
% with the output in form bb=[B-$... Bi ... B-1] and @=inv(Ho+H1*B-1)
|
||||||
% Dynare jacobian = [fy'-$... fy'i ... fy'+& fu']
|
% Dynare jacobian = [fy'-$... fy'i ... fy'+& fu']
|
||||||
% where [fy'-$... fy'i ... fy'+&]=[H-$... Hi ... H+&] and fu'= £
|
% where [fy'-$... fy'i ... fy'+&]=[H-$... Hi ... H+&] and fu'= £
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% jacobia_ [matrix] 1st order derivative of the model
|
% jacobia_ [matrix] 1st order derivative of the model
|
||||||
% dr [matlab structure] Decision rules for stochastic simulations.
|
% dr [matlab structure] Decision rules for stochastic simulations.
|
||||||
% M_ [matlab structure] Definition of the model.
|
% M_ [matlab structure] Definition of the model.
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% dr [matlab structure] Decision rules for stochastic simulations.
|
% dr [matlab structure] Decision rules for stochastic simulations.
|
||||||
% aimcode [integer] 1: the model defines variables uniquely
|
% aimcode [integer] 1: the model defines variables uniquely
|
||||||
|
@ -31,22 +31,22 @@ function [dr,aimcode,rts]=dynAIMsolver1(jacobia_,M_,dr)
|
||||||
% (c==63) e='Aim: A is NAN or INF.';
|
% (c==63) e='Aim: A is NAN or INF.';
|
||||||
% (c==64) e='Aim: Problem in SPEIG.';
|
% (c==64) e='Aim: Problem in SPEIG.';
|
||||||
% else e='Aimerr: return code not properly specified';
|
% else e='Aimerr: return code not properly specified';
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% Dynare use:
|
% Dynare use:
|
||||||
% 1) the lognormal block in DR1 is being invoked for some models and changing
|
% 1) the lognormal block in DR1 is being invoked for some models and changing
|
||||||
% values of ghx and ghy. We need to return the AIM output
|
% values of ghx and ghy. We need to return the AIM output
|
||||||
% values before that block and run the block with the current returned values
|
% values before that block and run the block with the current returned values
|
||||||
% of gy (i.e. dr.ghx) and gu (dr.ghu) if it is needed even when the AIM is used
|
% of gy (i.e. dr.ghx) and gu (dr.ghu) if it is needed even when the AIM is used
|
||||||
% (it does not depend on mjdgges output).
|
% (it does not depend on mjdgges output).
|
||||||
%
|
|
||||||
% 2) passing in aa={Q'|1}*jacobia_ can produce ~ one order closer
|
|
||||||
% results to the Dynare solutiion then when if plain jacobia_ is passed,
|
|
||||||
% i.e. diff < e-14 for aa and diff < *e-13 for jacobia_ if Q' is used.
|
|
||||||
%
|
%
|
||||||
% GP July 2008
|
% 2) passing in aa={Q'|1}*jacobia_ can produce ~ one order closer
|
||||||
|
% results to the Dynare solutiion then when if plain jacobia_ is passed,
|
||||||
|
% i.e. diff < e-14 for aa and diff < *e-13 for jacobia_ if Q' is used.
|
||||||
|
%
|
||||||
|
% GP July 2008
|
||||||
|
|
||||||
% Copyright (C) 2008-2012 Dynare Team
|
% Copyright (C) 2008-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -69,8 +69,8 @@ lags=M_.maximum_endo_lag; % no of lags and leads
|
||||||
leads=M_.maximum_endo_lead;
|
leads=M_.maximum_endo_lead;
|
||||||
klen=(leads+lags+1); % total lenght
|
klen=(leads+lags+1); % total lenght
|
||||||
theAIM_H=zeros(neq, neq*klen); % alloc space
|
theAIM_H=zeros(neq, neq*klen); % alloc space
|
||||||
lli=M_.lead_lag_incidence';
|
lli=M_.lead_lag_incidence';
|
||||||
% "sparse" the compact jacobia into AIM H aray of matrices
|
% "sparse" the compact jacobia into AIM H aray of matrices
|
||||||
% without exogenous shoc
|
% without exogenous shoc
|
||||||
theAIM_H(:,find(lli(:)))=jacobia_(:,nonzeros(lli(:)));
|
theAIM_H(:,find(lli(:)))=jacobia_(:,nonzeros(lli(:)));
|
||||||
condn = 1.e-10;%SPAmalg uses this in zero tests
|
condn = 1.e-10;%SPAmalg uses this in zero tests
|
||||||
|
@ -102,7 +102,7 @@ if aimcode==1 %if OK
|
||||||
col_order=[((i-1)*neq)+dr.order_var' col_order];
|
col_order=[((i-1)*neq)+dr.order_var' col_order];
|
||||||
end
|
end
|
||||||
bb_ord= bb(dr.order_var,col_order); % derive ordered gy
|
bb_ord= bb(dr.order_var,col_order); % derive ordered gy
|
||||||
|
|
||||||
% variables are present in the state space at the lag at which they
|
% variables are present in the state space at the lag at which they
|
||||||
% appear and at all smaller lags. The are ordered from smaller to
|
% appear and at all smaller lags. The are ordered from smaller to
|
||||||
% higher lag (reversed order of M_.lead_lag_incidence rows for lagged
|
% higher lag (reversed order of M_.lead_lag_incidence rows for lagged
|
||||||
|
@ -117,7 +117,7 @@ if aimcode==1 %if OK
|
||||||
%theH0= theAIM_H (:,lags*neq+1: (lags+1)*neq);
|
%theH0= theAIM_H (:,lags*neq+1: (lags+1)*neq);
|
||||||
% theHP= theAIM_H (:,(M_.maximum_endo_lag+1)*neq+1: (M_.maximum_endo_lag+2)*neq);
|
% theHP= theAIM_H (:,(M_.maximum_endo_lag+1)*neq+1: (M_.maximum_endo_lag+2)*neq);
|
||||||
%theHP= theAIM_H (:,(lags+1)*neq+1: (lags+2)*neq);
|
%theHP= theAIM_H (:,(lags+1)*neq+1: (lags+2)*neq);
|
||||||
theAIM_Psi= - jacobia_(:, size(nonzeros(lli(:)))+1:end);%
|
theAIM_Psi= - jacobia_(:, size(nonzeros(lli(:)))+1:end);%
|
||||||
%? = inv(H0 + H1B1)
|
%? = inv(H0 + H1B1)
|
||||||
%phi= (theH0+theHP*sparse(bb(:,(lags-1)*neq+1:end)))\eye( neq);
|
%phi= (theH0+theHP*sparse(bb(:,(lags-1)*neq+1:end)))\eye( neq);
|
||||||
%AIM_ghu=phi*theAIM_Psi;
|
%AIM_ghu=phi*theAIM_Psi;
|
||||||
|
@ -137,8 +137,8 @@ else
|
||||||
if aimcode < 1 || aimcode > 5 % too big exception, use mjdgges
|
if aimcode < 1 || aimcode > 5 % too big exception, use mjdgges
|
||||||
error('Error in AIM: aimcode=%d ; %s', aimcode, err);
|
error('Error in AIM: aimcode=%d ; %s', aimcode, err);
|
||||||
end
|
end
|
||||||
% if aimcode > 5
|
% if aimcode > 5
|
||||||
% disp(['Error in AIM: aimcode=' sprintf('%d : %s',aimcode, err)]);
|
% disp(['Error in AIM: aimcode=' sprintf('%d : %s',aimcode, err)]);
|
||||||
% aimcode=5;
|
% aimcode=5;
|
||||||
% end
|
% end
|
||||||
end
|
end
|
||||||
|
|
|
@ -51,7 +51,7 @@ function [dr,info]=AIM_first_order_solver(jacobia,M,dr,qz_criterium)
|
||||||
%! @end deftypefn
|
%! @end deftypefn
|
||||||
%@eod:
|
%@eod:
|
||||||
|
|
||||||
% Copyright (C) 2001-2016 Dynare Team
|
% Copyright (C) 2001-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -91,11 +91,10 @@ if nba ~= nsfwrd
|
||||||
if nba > nsfwrd
|
if nba > nsfwrd
|
||||||
temp = temp(nd-nba+1:nd-nsfwrd)-1-qz_criterium;
|
temp = temp(nd-nba+1:nd-nsfwrd)-1-qz_criterium;
|
||||||
info(1) = 3;
|
info(1) = 3;
|
||||||
elseif nba < nsfwrd;
|
elseif nba < nsfwrd
|
||||||
temp = temp(nd-nsfwrd+1:nd-nba)-1-qz_criterium;
|
temp = temp(nd-nsfwrd+1:nd-nba)-1-qz_criterium;
|
||||||
info(1) = 4;
|
info(1) = 4;
|
||||||
end
|
end
|
||||||
info(2) = temp'*temp;
|
info(2) = temp'*temp;
|
||||||
return
|
return
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
|
@ -2,7 +2,7 @@ function [DirectoryName, info] = CheckPath(type,dname)
|
||||||
% Creates the subfolder "./M_.dname/type" if it does not exist yet.
|
% Creates the subfolder "./M_.dname/type" if it does not exist yet.
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% type [string] Name of the subfolder.
|
% type [string] Name of the subfolder.
|
||||||
% dname [string] Name of the directory
|
% dname [string] Name of the directory
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
|
@ -12,7 +12,7 @@ function [DirectoryName, info] = CheckPath(type,dname)
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2005-2013 Dynare Team
|
% Copyright (C) 2005-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
|
|
@ -16,7 +16,7 @@ function CutSample(M_, options_, estim_params_)
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2005-2015 Dynare Team
|
% Copyright (C) 2005-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -66,7 +66,7 @@ if (TotalNumberOfMhFiles-1)-(FirstMhFile+1)+1 > 0
|
||||||
elseif TotalNumberOfMhFiles == 1
|
elseif TotalNumberOfMhFiles == 1
|
||||||
record.KeepedDraws.Distribution = [];
|
record.KeepedDraws.Distribution = [];
|
||||||
elseif TotalNumberOfMhFiles == 2 && FirstMhFile > 1
|
elseif TotalNumberOfMhFiles == 2 && FirstMhFile > 1
|
||||||
record.KeepedDraws.Distribution = [MAX_nruns-FirstLine+1 ; record.MhDraws(end,3)];
|
record.KeepedDraws.Distribution = [MAX_nruns-FirstLine+1 ; record.MhDraws(end,3)];
|
||||||
end
|
end
|
||||||
|
|
||||||
% Save updated mh-history file.
|
% Save updated mh-history file.
|
||||||
|
|
|
@ -1,8 +1,8 @@
|
||||||
function [alphahat,etahat,epsilonhat,ahat,SteadyState,trend_coeff,aK,T,R,P,PK,decomp,trend_addition,state_uncertainty,M_,oo_,options_,bayestopt_] = DsgeSmoother(xparam1,gend,Y,data_index,missing_value,M_,oo_,options_,bayestopt_,estim_params_)
|
function [alphahat,etahat,epsilonhat,ahat,SteadyState,trend_coeff,aK,T,R,P,PK,decomp,trend_addition,state_uncertainty,M_,oo_,options_,bayestopt_] = DsgeSmoother(xparam1,gend,Y,data_index,missing_value,M_,oo_,options_,bayestopt_,estim_params_)
|
||||||
% Estimation of the smoothed variables and innovations.
|
% Estimation of the smoothed variables and innovations.
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% o xparam1 [double] (p*1) vector of (estimated) parameters.
|
% o xparam1 [double] (p*1) vector of (estimated) parameters.
|
||||||
% o gend [integer] scalar specifying the number of observations ==> varargin{1}.
|
% o gend [integer] scalar specifying the number of observations ==> varargin{1}.
|
||||||
% o data [double] (n*T) matrix of data.
|
% o data [double] (n*T) matrix of data.
|
||||||
% o data_index [cell] 1*smpl cell of column vectors of indices.
|
% o data_index [cell] 1*smpl cell of column vectors of indices.
|
||||||
|
@ -12,7 +12,7 @@ function [alphahat,etahat,epsilonhat,ahat,SteadyState,trend_coeff,aK,T,R,P,PK,de
|
||||||
% o options_ [structure] describing the options
|
% o options_ [structure] describing the options
|
||||||
% o bayestopt_ [structure] describing the priors
|
% o bayestopt_ [structure] describing the priors
|
||||||
% o estim_params_ [structure] characterizing parameters to be estimated
|
% o estim_params_ [structure] characterizing parameters to be estimated
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% o alphahat [double] (m*T) matrix, smoothed endogenous variables (a_{t|T}) (decision-rule order)
|
% o alphahat [double] (m*T) matrix, smoothed endogenous variables (a_{t|T}) (decision-rule order)
|
||||||
% o etahat [double] (r*T) matrix, smoothed structural shocks (r>=n is the number of shocks).
|
% o etahat [double] (r*T) matrix, smoothed structural shocks (r>=n is the number of shocks).
|
||||||
|
@ -29,36 +29,36 @@ function [alphahat,etahat,epsilonhat,ahat,SteadyState,trend_coeff,aK,T,R,P,PK,de
|
||||||
% matrices (meaningless for periods 1:d) (decision-rule order)
|
% matrices (meaningless for periods 1:d) (decision-rule order)
|
||||||
% o decomp (K*m*r*(T+K)) 4D array of shock decomposition of k-step ahead
|
% o decomp (K*m*r*(T+K)) 4D array of shock decomposition of k-step ahead
|
||||||
% filtered variables (decision-rule order)
|
% filtered variables (decision-rule order)
|
||||||
% o trend_addition [double] (n*T) pure trend component; stored in options_.varobs order
|
% o trend_addition [double] (n*T) pure trend component; stored in options_.varobs order
|
||||||
% o state_uncertainty [double] (K,K,T) array, storing the uncertainty
|
% o state_uncertainty [double] (K,K,T) array, storing the uncertainty
|
||||||
% about the smoothed state (decision-rule order)
|
% about the smoothed state (decision-rule order)
|
||||||
% o M_ [structure] decribing the model
|
% o M_ [structure] decribing the model
|
||||||
% o oo_ [structure] storing the results
|
% o oo_ [structure] storing the results
|
||||||
% o options_ [structure] describing the options
|
% o options_ [structure] describing the options
|
||||||
% o bayestopt_ [structure] describing the priors
|
% o bayestopt_ [structure] describing the priors
|
||||||
%
|
%
|
||||||
% Notes:
|
% Notes:
|
||||||
% m: number of endogenous variables (M_.endo_nbr)
|
% m: number of endogenous variables (M_.endo_nbr)
|
||||||
% T: number of Time periods (options_.nobs)
|
% T: number of Time periods (options_.nobs)
|
||||||
% r: number of strucural shocks (M_.exo_nbr)
|
% r: number of strucural shocks (M_.exo_nbr)
|
||||||
% n: number of observables (length(options_.varobs))
|
% n: number of observables (length(options_.varobs))
|
||||||
% K: maximum forecast horizon (max(options_.nk))
|
% K: maximum forecast horizon (max(options_.nk))
|
||||||
%
|
%
|
||||||
% To get variables that are stored in decision rule order in order of declaration
|
% To get variables that are stored in decision rule order in order of declaration
|
||||||
% as in M_.endo_names, ones needs code along the lines of:
|
% as in M_.endo_names, ones needs code along the lines of:
|
||||||
% variables_declaration_order(dr.order_var,:) = alphahat
|
% variables_declaration_order(dr.order_var,:) = alphahat
|
||||||
%
|
%
|
||||||
% Defines bayestopt_.mf = bayestopt_.smoother_mf (positions of observed variables
|
% Defines bayestopt_.mf = bayestopt_.smoother_mf (positions of observed variables
|
||||||
% and requested smoothed variables in decision rules (decision rule order)) and
|
% and requested smoothed variables in decision rules (decision rule order)) and
|
||||||
% passes it back via global variable
|
% passes it back via global variable
|
||||||
%
|
%
|
||||||
% ALGORITHM
|
% ALGORITHM
|
||||||
% Diffuse Kalman filter (Durbin and Koopman)
|
% Diffuse Kalman filter (Durbin and Koopman)
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% None
|
% None
|
||||||
|
|
||||||
% Copyright (C) 2006-2016 Dynare Team
|
% Copyright (C) 2006-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -97,7 +97,7 @@ end
|
||||||
%------------------------------------------------------------------------------
|
%------------------------------------------------------------------------------
|
||||||
% 2. call model setup & reduction program
|
% 2. call model setup & reduction program
|
||||||
%------------------------------------------------------------------------------
|
%------------------------------------------------------------------------------
|
||||||
oldoo.restrict_var_list = oo_.dr.restrict_var_list;
|
oldoo.restrict_var_list = oo_.dr.restrict_var_list;
|
||||||
oldoo.restrict_columns = oo_.dr.restrict_columns;
|
oldoo.restrict_columns = oo_.dr.restrict_columns;
|
||||||
oo_.dr.restrict_var_list = bayestopt_.smoother_var_list;
|
oo_.dr.restrict_var_list = bayestopt_.smoother_var_list;
|
||||||
oo_.dr.restrict_columns = bayestopt_.smoother_restrict_columns;
|
oo_.dr.restrict_columns = bayestopt_.smoother_restrict_columns;
|
||||||
|
@ -133,8 +133,8 @@ mf = bayestopt_.mf;
|
||||||
% ------------------------------------------------------------------------------
|
% ------------------------------------------------------------------------------
|
||||||
% 3. Initial condition of the Kalman filter
|
% 3. Initial condition of the Kalman filter
|
||||||
% ------------------------------------------------------------------------------
|
% ------------------------------------------------------------------------------
|
||||||
%
|
%
|
||||||
% Here, Pinf and Pstar are determined. If the model is stationary, determine
|
% Here, Pinf and Pstar are determined. If the model is stationary, determine
|
||||||
% Pstar as the solution of the Lyapunov equation and set Pinf=[] (Notation follows
|
% Pstar as the solution of the Lyapunov equation and set Pinf=[] (Notation follows
|
||||||
% Koopman/Durbin (2003), Journal of Time Series Analysis 24(1))
|
% Koopman/Durbin (2003), Journal of Time Series Analysis 24(1))
|
||||||
%
|
%
|
||||||
|
@ -256,7 +256,7 @@ if kalman_algo == 2 || kalman_algo == 4
|
||||||
else
|
else
|
||||||
Pstar = blkdiag(Pstar,H);
|
Pstar = blkdiag(Pstar,H);
|
||||||
if ~isempty(Pinf)
|
if ~isempty(Pinf)
|
||||||
Pinf = blkdiag(Pinf,zeros(vobs));
|
Pinf = blkdiag(Pinf,zeros(vobs));
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
%now reset H to 0
|
%now reset H to 0
|
||||||
|
@ -265,7 +265,7 @@ if kalman_algo == 2 || kalman_algo == 4
|
||||||
%do nothing, state vector was already expanded
|
%do nothing, state vector was already expanded
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
[alphahat,epsilonhat,etahat,ahat,P,aK,PK,decomp,state_uncertainty] = missing_DiffuseKalmanSmootherH3_Z(ST, ...
|
[alphahat,epsilonhat,etahat,ahat,P,aK,PK,decomp,state_uncertainty] = missing_DiffuseKalmanSmootherH3_Z(ST, ...
|
||||||
Z,R1,Q,diag(H), ...
|
Z,R1,Q,diag(H), ...
|
||||||
Pinf,Pstar,data1,vobs,np,smpl,data_index, ...
|
Pinf,Pstar,data1,vobs,np,smpl,data_index, ...
|
||||||
|
@ -282,7 +282,7 @@ if expanded_state_vector_for_univariate_filter && (kalman_algo == 2 || kalman_al
|
||||||
ahat = ahat(k,:);
|
ahat = ahat(k,:);
|
||||||
aK = aK(:,k,:,:);
|
aK = aK(:,k,:,:);
|
||||||
epsilonhat=etahat(end-vobs+1:end,:);
|
epsilonhat=etahat(end-vobs+1:end,:);
|
||||||
etahat=etahat(1:end-vobs,:);
|
etahat=etahat(1:end-vobs,:);
|
||||||
if ~isempty(PK)
|
if ~isempty(PK)
|
||||||
PK = PK(:,k,k,:);
|
PK = PK(:,k,k,:);
|
||||||
end
|
end
|
||||||
|
|
|
@ -5,18 +5,18 @@ function Draws = GetAllPosteriorDraws(column, FirstMhFile, FirstLine, TotalNumbe
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% column: column
|
% column: column
|
||||||
% FirstMhFile: first mh file
|
% FirstMhFile: first mh file
|
||||||
% FirstLine: first line
|
% FirstLine: first line
|
||||||
% TotalNumberOfMhFile: total number of mh file
|
% TotalNumberOfMhFile: total number of mh file
|
||||||
% NumberOfDraws: number of draws
|
% NumberOfDraws: number of draws
|
||||||
|
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% Draws: draws from posterior distribution
|
% Draws: draws from posterior distribution
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2005-2011 Dynare Team
|
% Copyright (C) 2005-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -55,7 +55,7 @@ if nblck>1 && nargin<6
|
||||||
iline = 1;
|
iline = 1;
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
else
|
else
|
||||||
for blck = 1:nblck
|
for blck = 1:nblck
|
||||||
iline=iline0;
|
iline=iline0;
|
||||||
for file = FirstMhFile:TotalNumberOfMhFile
|
for file = FirstMhFile:TotalNumberOfMhFile
|
||||||
|
|
|
@ -5,15 +5,15 @@ function [xparams, logpost] = GetOneDraw(type)
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% type: [string] 'posterior': draw from MCMC draws
|
% type: [string] 'posterior': draw from MCMC draws
|
||||||
% 'prior': draw from prior
|
% 'prior': draw from prior
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% xparams: vector of estimated parameters (drawn from posterior or prior distribution)
|
% xparams: vector of estimated parameters (drawn from posterior or prior distribution)
|
||||||
% logpost: log of the posterior density of this parameter vector
|
% logpost: log of the posterior density of this parameter vector
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2005-2015 Dynare Team
|
% Copyright (C) 2005-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
|
|
@ -1,6 +1,6 @@
|
||||||
function [mean,variance] = GetPosteriorMeanVariance(M,drop)
|
function [mean,variance] = GetPosteriorMeanVariance(M,drop)
|
||||||
|
|
||||||
% Copyright (C) 2012-2016 Dynare Team
|
% Copyright (C) 2012-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
|
|
@ -1,22 +1,22 @@
|
||||||
function oo_ = GetPosteriorParametersStatistics(estim_params_, M_, options_, bayestopt_, oo_, pnames)
|
function oo_ = GetPosteriorParametersStatistics(estim_params_, M_, options_, bayestopt_, oo_, pnames)
|
||||||
% This function prints and saves posterior estimates after the mcmc
|
% This function prints and saves posterior estimates after the mcmc
|
||||||
% (+updates of oo_ & TeX output).
|
% (+updates of oo_ & TeX output).
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% estim_params_ [structure]
|
% estim_params_ [structure]
|
||||||
% M_ [structure]
|
% M_ [structure]
|
||||||
% options_ [structure]
|
% options_ [structure]
|
||||||
% bayestopt_ [structure]
|
% bayestopt_ [structure]
|
||||||
% oo_ [structure]
|
% oo_ [structure]
|
||||||
% pnames [char] Array of char, names of the prior shapes available
|
% pnames [char] Array of char, names of the prior shapes available
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% oo_ [structure]
|
% oo_ [structure]
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% None.
|
% None.
|
||||||
|
|
||||||
% Copyright (C) 2006-2016 Dynare Team
|
% Copyright (C) 2006-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -34,7 +34,7 @@ function oo_ = GetPosteriorParametersStatistics(estim_params_, M_, options_, bay
|
||||||
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
|
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
|
||||||
|
|
||||||
%if ~options_.mh_replic && options_.load_mh_file
|
%if ~options_.mh_replic && options_.load_mh_file
|
||||||
% load([M_.fname '_results.mat'],'oo_');
|
% load([M_.fname '_results.mat'],'oo_');
|
||||||
%end
|
%end
|
||||||
|
|
||||||
TeX = options_.TeX;
|
TeX = options_.TeX;
|
||||||
|
@ -48,7 +48,7 @@ nx = nvx+nvn+ncx+ncn+np;
|
||||||
|
|
||||||
MetropolisFolder = CheckPath('metropolis',M_.dname);
|
MetropolisFolder = CheckPath('metropolis',M_.dname);
|
||||||
OutputFolder = CheckPath('Output',M_.dname);
|
OutputFolder = CheckPath('Output',M_.dname);
|
||||||
FileName = M_.fname;
|
FileName = M_.fname;
|
||||||
|
|
||||||
load_last_mh_history_file(MetropolisFolder,FileName);
|
load_last_mh_history_file(MetropolisFolder,FileName);
|
||||||
|
|
||||||
|
@ -108,14 +108,14 @@ if np
|
||||||
[post_mean, post_median, post_var, hpd_interval, post_deciles, ...
|
[post_mean, post_median, post_var, hpd_interval, post_deciles, ...
|
||||||
density] = posterior_moments(Draws,1,options_.mh_conf_sig);
|
density] = posterior_moments(Draws,1,options_.mh_conf_sig);
|
||||||
name = bayestopt_.name{ip};
|
name = bayestopt_.name{ip};
|
||||||
oo_ = Filloo(oo_,name,type,post_mean,hpd_interval,post_median,post_var,post_deciles,density);
|
oo_ = Filloo(oo_,name,type,post_mean,hpd_interval,post_median,post_var,post_deciles,density);
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
disp(sprintf(pformat,header_width,name,bayestopt_.p1(ip),...
|
disp(sprintf(pformat,header_width,name,bayestopt_.p1(ip),...
|
||||||
post_mean, ...
|
post_mean, ...
|
||||||
hpd_interval, ...
|
hpd_interval, ...
|
||||||
pnames(bayestopt_.pshape(ip)+1,:), ...
|
pnames(bayestopt_.pshape(ip)+1,:), ...
|
||||||
bayestopt_.p2(ip)));
|
bayestopt_.p2(ip)));
|
||||||
if TeX
|
if TeX
|
||||||
k = estim_params_.param_vals(i,1);
|
k = estim_params_.param_vals(i,1);
|
||||||
name = deblank(M_.param_names_tex(k,:));
|
name = deblank(M_.param_names_tex(k,:));
|
||||||
|
@ -163,7 +163,7 @@ if nvx
|
||||||
end
|
end
|
||||||
disp(sprintf(pformat,header_width,name,bayestopt_.p1(ip),post_mean,hpd_interval,...
|
disp(sprintf(pformat,header_width,name,bayestopt_.p1(ip),post_mean,hpd_interval,...
|
||||||
pnames(bayestopt_.pshape(ip)+1,:),bayestopt_.p2(ip)));
|
pnames(bayestopt_.pshape(ip)+1,:),bayestopt_.p2(ip)));
|
||||||
if TeX,
|
if TeX
|
||||||
name = deblank(M_.exo_names_tex(k,:));
|
name = deblank(M_.exo_names_tex(k,:));
|
||||||
TeXCore(fid,name,deblank(pnames(bayestopt_.pshape(ip)+1,:)),bayestopt_.p1(ip),...
|
TeXCore(fid,name,deblank(pnames(bayestopt_.pshape(ip)+1,:)),bayestopt_.p1(ip),...
|
||||||
bayestopt_.p2(ip),post_mean,sqrt(post_var),hpd_interval);
|
bayestopt_.p2(ip),post_mean,sqrt(post_var),hpd_interval);
|
||||||
|
@ -171,7 +171,7 @@ if nvx
|
||||||
ip = ip+1;
|
ip = ip+1;
|
||||||
end
|
end
|
||||||
if TeX
|
if TeX
|
||||||
TeXEnd(fid,2,'standard deviation of structural shocks');
|
TeXEnd(fid,2,'standard deviation of structural shocks');
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
if nvn
|
if nvn
|
||||||
|
@ -213,7 +213,7 @@ if nvn
|
||||||
ip = ip+1;
|
ip = ip+1;
|
||||||
end
|
end
|
||||||
if TeX
|
if TeX
|
||||||
TeXEnd(fid,3,'standard deviation of measurement errors');
|
TeXEnd(fid,3,'standard deviation of measurement errors');
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
if ncx
|
if ncx
|
||||||
|
@ -311,15 +311,15 @@ if ncn
|
||||||
end
|
end
|
||||||
disp(sprintf(pformat, header_width,name,bayestopt_.p1(ip),post_mean,hpd_interval, ...
|
disp(sprintf(pformat, header_width,name,bayestopt_.p1(ip),post_mean,hpd_interval, ...
|
||||||
pnames(bayestopt_.pshape(ip)+1,:),bayestopt_.p2(ip)));
|
pnames(bayestopt_.pshape(ip)+1,:),bayestopt_.p2(ip)));
|
||||||
if TeX,
|
if TeX
|
||||||
name = ['(',deblank(M_.endo_names_tex(k1,:)) ',' deblank(M_.endo_names_tex(k2,:)),')'];
|
name = ['(',deblank(M_.endo_names_tex(k1,:)) ',' deblank(M_.endo_names_tex(k2,:)),')'];
|
||||||
TeXCore(fid,name,deblank(pnames(bayestopt_.pshape(ip)+1,:)),bayestopt_.p1(ip),...
|
TeXCore(fid,name,deblank(pnames(bayestopt_.pshape(ip)+1,:)),bayestopt_.p1(ip),...
|
||||||
bayestopt_.p2(ip),post_mean,sqrt(post_var),hpd_interval);
|
bayestopt_.p2(ip),post_mean,sqrt(post_var),hpd_interval);
|
||||||
end
|
end
|
||||||
ip = ip+1;
|
ip = ip+1;
|
||||||
end
|
end
|
||||||
if TeX
|
if TeX
|
||||||
TeXEnd(fid,5,'correlation of measurement errors');
|
TeXEnd(fid,5,'correlation of measurement errors');
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
|
@ -359,7 +359,7 @@ fid = fidTeX;
|
||||||
|
|
||||||
|
|
||||||
function TeXCore(fid,name,shape,priormean,priorstd,postmean,poststd,hpd)
|
function TeXCore(fid,name,shape,priormean,priorstd,postmean,poststd,hpd)
|
||||||
fprintf(fid,['$%s$ & %s & %7.3f & %6.4f & %7.3f& %6.4f & %7.4f & %7.4f \\\\ \n'],...
|
fprintf(fid,['$%s$ & %s & %7.3f & %6.4f & %7.3f& %6.4f & %7.4f & %7.4f \\\\ \n'],...
|
||||||
name,...
|
name,...
|
||||||
shape,...
|
shape,...
|
||||||
priormean,...
|
priormean,...
|
||||||
|
@ -371,7 +371,7 @@ fprintf(fid,['$%s$ & %s & %7.3f & %6.4f & %7.3f& %6.4f & %7.4f & %7.4f \\\\ \n']
|
||||||
|
|
||||||
|
|
||||||
function TeXEnd(fid,fnum,title)
|
function TeXEnd(fid,fnum,title)
|
||||||
fprintf(fid,'\\end{longtable}\n ');
|
fprintf(fid,'\\end{longtable}\n ');
|
||||||
fprintf(fid,'\\end{center}\n');
|
fprintf(fid,'\\end{center}\n');
|
||||||
fprintf(fid,'%% End of TeX file.\n');
|
fprintf(fid,'%% End of TeX file.\n');
|
||||||
fclose(fid);
|
fclose(fid);
|
||||||
|
|
|
@ -1,6 +1,6 @@
|
||||||
function MakeAllFigures(NumberOfPlots,Caption,FigureProperties,Info)
|
function MakeAllFigures(NumberOfPlots,Caption,FigureProperties,Info)
|
||||||
|
|
||||||
% Copyright (C) 2005-2009 Dynare Team
|
% Copyright (C) 2005-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -19,11 +19,11 @@ function MakeAllFigures(NumberOfPlots,Caption,FigureProperties,Info)
|
||||||
|
|
||||||
global M_ options_
|
global M_ options_
|
||||||
|
|
||||||
FigHandle = figure('Name',FigureProperties.Name);
|
FigHandle = figure('Name',FigureProperties.Name);
|
||||||
|
|
||||||
NAMES = cell(NumberOfPlots,1);
|
NAMES = cell(NumberOfPlots,1);
|
||||||
if options_.TeX
|
if options_.TeX
|
||||||
TeXNAMES = cell(NumberOfPlots,1);
|
TeXNAMES = cell(NumberOfPlots,1);
|
||||||
end
|
end
|
||||||
|
|
||||||
if NumberOfPlots == 9
|
if NumberOfPlots == 9
|
||||||
|
@ -53,7 +53,7 @@ elseif NumberOfPlots == 2
|
||||||
elseif NumberOfPlots == 1
|
elseif NumberOfPlots == 1
|
||||||
nr = 1;
|
nr = 1;
|
||||||
nc = 1;
|
nc = 1;
|
||||||
end
|
end
|
||||||
|
|
||||||
for plt = 1:NumberOfPlots
|
for plt = 1:NumberOfPlots
|
||||||
eval(['NumberOfCurves = Info.Box' int2str(plt) '.Number;'])
|
eval(['NumberOfCurves = Info.Box' int2str(plt) '.Number;'])
|
||||||
|
@ -138,7 +138,7 @@ for plt = 1:NumberOfPlots
|
||||||
set(hh,'Color','r','LineStyle','-','LineWidth',2)
|
set(hh,'Color','r','LineStyle','-','LineWidth',2)
|
||||||
%
|
%
|
||||||
%
|
%
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
axis([xmin xmax ymin ymax])
|
axis([xmin xmax ymin ymax])
|
||||||
title(NAMES{plt})
|
title(NAMES{plt})
|
||||||
|
@ -150,14 +150,14 @@ if Info.SaveFormat.Eps
|
||||||
if isempty(Info.SaveFormat.Name)
|
if isempty(Info.SaveFormat.Name)
|
||||||
eval(['print -depsc2 ' M_.fname Info.SaveFormat.GenericName int2str(Info.SaveFormat.Number) '.eps']);
|
eval(['print -depsc2 ' M_.fname Info.SaveFormat.GenericName int2str(Info.SaveFormat.Number) '.eps']);
|
||||||
else
|
else
|
||||||
eval(['print -depsc2 ' M_.fname Info.SaveFormat.GenericName Info.SaveFormat.Name '.eps']);
|
eval(['print -depsc2 ' M_.fname Info.SaveFormat.GenericName Info.SaveFormat.Name '.eps']);
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
if Info.SaveFormat.Pdf && ~isoctave
|
if Info.SaveFormat.Pdf && ~isoctave
|
||||||
if isempty(Info.SaveFormat.Name)
|
if isempty(Info.SaveFormat.Name)
|
||||||
eval(['print -dpdf ' M_.fname Info.SaveFormat.GenericName int2str(Info.SaveFormat.Number)]);
|
eval(['print -dpdf ' M_.fname Info.SaveFormat.GenericName int2str(Info.SaveFormat.Number)]);
|
||||||
else
|
else
|
||||||
eval(['print -dpdf ' M_.fname Info.SaveFormat.GenericName Info.SaveFormat.Name]);
|
eval(['print -dpdf ' M_.fname Info.SaveFormat.GenericName Info.SaveFormat.Name]);
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
if Info.SaveFormat.Fig && ~isoctave
|
if Info.SaveFormat.Fig && ~isoctave
|
||||||
|
|
|
@ -4,15 +4,15 @@ function oo_ = PlotPosteriorDistributions(estim_params_, M_, options_, bayestopt
|
||||||
% plots posterior distributions
|
% plots posterior distributions
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% estim_params_ [structure]
|
% estim_params_ [structure]
|
||||||
% M_ [structure]
|
% M_ [structure]
|
||||||
% options_ [structure]
|
% options_ [structure]
|
||||||
% bayestopt_ [structure]
|
% bayestopt_ [structure]
|
||||||
% oo_ [structure]
|
% oo_ [structure]
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% oo_ [structure]
|
% oo_ [structure]
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
|
@ -49,7 +49,7 @@ nn = sqrt(MaxNumberOfPlotPerFigure);
|
||||||
|
|
||||||
figurename = 'Priors and posteriors';
|
figurename = 'Priors and posteriors';
|
||||||
|
|
||||||
if TeX && any(strcmp('eps',cellstr(options_.graph_format)))
|
if TeX && any(strcmp('eps',cellstr(options_.graph_format)))
|
||||||
fidTeX = fopen([OutputDirectoryName '/' M_.fname '_PriorsAndPosteriors.tex'],'w');
|
fidTeX = fopen([OutputDirectoryName '/' M_.fname '_PriorsAndPosteriors.tex'],'w');
|
||||||
fprintf(fidTeX,'%% TeX eps-loader file generated by PlotPosteriorDistributions.m (Dynare).\n');
|
fprintf(fidTeX,'%% TeX eps-loader file generated by PlotPosteriorDistributions.m (Dynare).\n');
|
||||||
fprintf(fidTeX,['%% ' datestr(now,0) '\n']);
|
fprintf(fidTeX,['%% ' datestr(now,0) '\n']);
|
||||||
|
@ -78,9 +78,9 @@ for i=1:npar
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
[x2,f2,abscissa,dens,binf2,bsup2] = draw_prior_density(i,bayestopt_);
|
[x2,f2,abscissa,dens,binf2,bsup2] = draw_prior_density(i,bayestopt_);
|
||||||
top2 = max(f2);
|
top2 = max(f2);
|
||||||
if i <= nvx
|
if i <= nvx
|
||||||
name = deblank(M_.exo_names(estim_params_.var_exo(i,1),:));
|
name = deblank(M_.exo_names(estim_params_.var_exo(i,1),:));
|
||||||
x1 = oo_.posterior_density.shocks_std.(name)(:,1);
|
x1 = oo_.posterior_density.shocks_std.(name)(:,1);
|
||||||
f1 = oo_.posterior_density.shocks_std.(name)(:,2);
|
f1 = oo_.posterior_density.shocks_std.(name)(:,2);
|
||||||
oo_.prior_density.shocks_std.(name)(:,1) = x2;
|
oo_.prior_density.shocks_std.(name)(:,1) = x2;
|
||||||
|
@ -96,18 +96,18 @@ for i=1:npar
|
||||||
oo_.prior_density.measurement_errors_std.(name)(:,2) = f2;
|
oo_.prior_density.measurement_errors_std.(name)(:,2) = f2;
|
||||||
if ~options_.mh_posterior_mode_estimation
|
if ~options_.mh_posterior_mode_estimation
|
||||||
pmod = oo_.posterior_mode.measurement_errors_std.(name);
|
pmod = oo_.posterior_mode.measurement_errors_std.(name);
|
||||||
end
|
end
|
||||||
elseif i <= nvx+nvn+ncx
|
elseif i <= nvx+nvn+ncx
|
||||||
j = i - (nvx+nvn);
|
j = i - (nvx+nvn);
|
||||||
k1 = estim_params_.corrx(j,1);
|
k1 = estim_params_.corrx(j,1);
|
||||||
k2 = estim_params_.corrx(j,2);
|
k2 = estim_params_.corrx(j,2);
|
||||||
name = [deblank(M_.exo_names(k1,:)) '_' deblank(M_.exo_names(k2,:))];
|
name = [deblank(M_.exo_names(k1,:)) '_' deblank(M_.exo_names(k2,:))];
|
||||||
x1 = oo_.posterior_density.shocks_corr.(name)(:,1);
|
x1 = oo_.posterior_density.shocks_corr.(name)(:,1);
|
||||||
f1 = oo_.posterior_density.shocks_corr.(name)(:,2);
|
f1 = oo_.posterior_density.shocks_corr.(name)(:,2);
|
||||||
oo_.prior_density.shocks_corr.(name)(:,1) = x2;
|
oo_.prior_density.shocks_corr.(name)(:,1) = x2;
|
||||||
oo_.prior_density.shocks_corr.(name)(:,2) = f2;
|
oo_.prior_density.shocks_corr.(name)(:,2) = f2;
|
||||||
if ~options_.mh_posterior_mode_estimation
|
if ~options_.mh_posterior_mode_estimation
|
||||||
pmod = oo_.posterior_mode.shocks_corr.(name);
|
pmod = oo_.posterior_mode.shocks_corr.(name);
|
||||||
end
|
end
|
||||||
elseif i <= nvx+nvn+ncx+ncn
|
elseif i <= nvx+nvn+ncx+ncn
|
||||||
j = i - (nvx+nvn+ncx);
|
j = i - (nvx+nvn+ncx);
|
||||||
|
@ -151,13 +151,13 @@ for i=1:npar
|
||||||
title(nam,'Interpreter','none');
|
title(nam,'Interpreter','none');
|
||||||
hold off;
|
hold off;
|
||||||
drawnow
|
drawnow
|
||||||
if subplotnum == MaxNumberOfPlotPerFigure || i == npar;
|
if subplotnum == MaxNumberOfPlotPerFigure || i == npar
|
||||||
dyn_saveas(hfig,[OutputDirectoryName '/' M_.fname '_PriorsAndPosteriors' int2str(figunumber)],options_.nodisplay,options_.graph_format);
|
dyn_saveas(hfig,[OutputDirectoryName '/' M_.fname '_PriorsAndPosteriors' int2str(figunumber)],options_.nodisplay,options_.graph_format);
|
||||||
if TeX && any(strcmp('eps',cellstr(options_.graph_format)))
|
if TeX && any(strcmp('eps',cellstr(options_.graph_format)))
|
||||||
fprintf(fidTeX,'\\begin{figure}[H]\n');
|
fprintf(fidTeX,'\\begin{figure}[H]\n');
|
||||||
for j = 1:size(NAMES,1)
|
for j = 1:size(NAMES,1)
|
||||||
fprintf(fidTeX,'\\psfrag{%s}[1][][0.5][0]{%s}\n',deblank(NAMES(j,:)),deblank(TeXNAMES(j,:)));
|
fprintf(fidTeX,'\\psfrag{%s}[1][][0.5][0]{%s}\n',deblank(NAMES(j,:)),deblank(TeXNAMES(j,:)));
|
||||||
end
|
end
|
||||||
fprintf(fidTeX,'\\centering\n');
|
fprintf(fidTeX,'\\centering\n');
|
||||||
fprintf(fidTeX,'\\includegraphics[width=%2.2f\\textwidth]{%s/%s_PriorsAndPosteriors%s}\n',options_.figures.textwidth*min(subplotnum/nn,1),OutputDirectoryName,M_.fname,int2str(figunumber));
|
fprintf(fidTeX,'\\includegraphics[width=%2.2f\\textwidth]{%s/%s_PriorsAndPosteriors%s}\n',options_.figures.textwidth*min(subplotnum/nn,1),OutputDirectoryName,M_.fname,int2str(figunumber));
|
||||||
fprintf(fidTeX,'\\caption{Priors and posteriors.}');
|
fprintf(fidTeX,'\\caption{Priors and posteriors.}');
|
||||||
|
|
|
@ -16,7 +16,7 @@ function PosteriorIRF(type)
|
||||||
% functions associated with it(the _core1 and _core2).
|
% functions associated with it(the _core1 and _core2).
|
||||||
% See also the comments posterior_sampler.m funtion.
|
% See also the comments posterior_sampler.m funtion.
|
||||||
|
|
||||||
% Copyright (C) 2006-2016 Dynare Team
|
% Copyright (C) 2006-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -178,7 +178,7 @@ if strcmpi(type,'posterior')
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
if ~strcmpi(type,'prior'),
|
if ~strcmpi(type,'prior')
|
||||||
localVars.x=x;
|
localVars.x=x;
|
||||||
end
|
end
|
||||||
|
|
||||||
|
@ -202,16 +202,16 @@ localVars.ifil2=ifil2;
|
||||||
localVars.MhDirectoryName=MhDirectoryName;
|
localVars.MhDirectoryName=MhDirectoryName;
|
||||||
|
|
||||||
% Like sequential execution!
|
% Like sequential execution!
|
||||||
if isnumeric(options_.parallel),
|
if isnumeric(options_.parallel)
|
||||||
[fout] = PosteriorIRF_core1(localVars,1,B,0);
|
[fout] = PosteriorIRF_core1(localVars,1,B,0);
|
||||||
nosaddle = fout.nosaddle;
|
nosaddle = fout.nosaddle;
|
||||||
else
|
else
|
||||||
% Parallel execution!
|
% Parallel execution!
|
||||||
[nCPU, totCPU, nBlockPerCPU] = distributeJobs(options_.parallel, 1, B);
|
[nCPU, totCPU, nBlockPerCPU] = distributeJobs(options_.parallel, 1, B);
|
||||||
for j=1:totCPU-1,
|
for j=1:totCPU-1
|
||||||
nfiles = ceil(nBlockPerCPU(j)/MAX_nirfs_dsge);
|
nfiles = ceil(nBlockPerCPU(j)/MAX_nirfs_dsge);
|
||||||
NumberOfIRFfiles_dsge(j+1) =NumberOfIRFfiles_dsge(j)+nfiles;
|
NumberOfIRFfiles_dsge(j+1) =NumberOfIRFfiles_dsge(j)+nfiles;
|
||||||
if MAX_nirfs_dsgevar,
|
if MAX_nirfs_dsgevar
|
||||||
nfiles = ceil(nBlockPerCPU(j)/MAX_nirfs_dsgevar);
|
nfiles = ceil(nBlockPerCPU(j)/MAX_nirfs_dsgevar);
|
||||||
else
|
else
|
||||||
nfiles=0;
|
nfiles=0;
|
||||||
|
@ -236,8 +236,8 @@ else
|
||||||
NamFileInput(1,:) = {'',[M_.fname '_static.m']};
|
NamFileInput(1,:) = {'',[M_.fname '_static.m']};
|
||||||
NamFileInput(2,:) = {'',[M_.fname '_dynamic.m']};
|
NamFileInput(2,:) = {'',[M_.fname '_dynamic.m']};
|
||||||
NamFileInput(3,:) = {'',[M_.fname '_set_auxiliary_variables.m']};
|
NamFileInput(3,:) = {'',[M_.fname '_set_auxiliary_variables.m']};
|
||||||
if options_.steadystate_flag,
|
if options_.steadystate_flag
|
||||||
if options_.steadystate_flag == 1,
|
if options_.steadystate_flag == 1
|
||||||
NamFileInput(length(NamFileInput)+1,:)={'',[M_.fname '_steadystate.m']};
|
NamFileInput(length(NamFileInput)+1,:)={'',[M_.fname '_steadystate.m']};
|
||||||
else
|
else
|
||||||
NamFileInput(length(NamFileInput)+1,:)={'',[M_.fname '_steadystate2.m']};
|
NamFileInput(length(NamFileInput)+1,:)={'',[M_.fname '_steadystate2.m']};
|
||||||
|
@ -245,7 +245,7 @@ else
|
||||||
end
|
end
|
||||||
[fout] = masterParallel(options_.parallel, 1, B,NamFileInput,'PosteriorIRF_core1', localVars, globalVars, options_.parallel_info);
|
[fout] = masterParallel(options_.parallel, 1, B,NamFileInput,'PosteriorIRF_core1', localVars, globalVars, options_.parallel_info);
|
||||||
nosaddle=0;
|
nosaddle=0;
|
||||||
for j=1:length(fout),
|
for j=1:length(fout)
|
||||||
nosaddle = nosaddle + fout(j).nosaddle;
|
nosaddle = nosaddle + fout(j).nosaddle;
|
||||||
end
|
end
|
||||||
|
|
||||||
|
@ -408,9 +408,9 @@ if ~options_.nograph && ~options_.no_graph.posterior
|
||||||
for jj=1:nvar
|
for jj=1:nvar
|
||||||
if max(abs(MeanIRF(:,jj,ii))) >= options_.impulse_responses.plot_threshold
|
if max(abs(MeanIRF(:,jj,ii))) >= options_.impulse_responses.plot_threshold
|
||||||
subplotnum = subplotnum+1;
|
subplotnum = subplotnum+1;
|
||||||
|
|
||||||
if subplotnum == 1
|
if subplotnum == 1
|
||||||
fprintf(fidTeX,'\\begin{figure}[H]\n');
|
fprintf(fidTeX,'\\begin{figure}[H]\n');
|
||||||
end
|
end
|
||||||
name = deblank(varlist(jj,:));
|
name = deblank(varlist(jj,:));
|
||||||
texname = deblank(varlist_TeX(jj,:));
|
texname = deblank(varlist_TeX(jj,:));
|
||||||
|
@ -418,7 +418,7 @@ if ~options_.nograph && ~options_.no_graph.posterior
|
||||||
end
|
end
|
||||||
if subplotnum == MaxNumberOfPlotPerFigure || (jj == nvar && subplotnum> 0)
|
if subplotnum == MaxNumberOfPlotPerFigure || (jj == nvar && subplotnum> 0)
|
||||||
figunumber = figunumber+1;
|
figunumber = figunumber+1;
|
||||||
|
|
||||||
fprintf(fidTeX,'\\centering \n');
|
fprintf(fidTeX,'\\centering \n');
|
||||||
fprintf(fidTeX,'\\includegraphics[width=%2.2f\\textwidth]{%s/%s_Bayesian_IRF_%s_%d}\n',options_.figures.textwidth*min(subplotnum/nn,1),DirectoryName,M_.fname,deblank(tit(ii,:)),figunumber);
|
fprintf(fidTeX,'\\includegraphics[width=%2.2f\\textwidth]{%s/%s_Bayesian_IRF_%s_%d}\n',options_.figures.textwidth*min(subplotnum/nn,1),DirectoryName,M_.fname,deblank(tit(ii,:)),figunumber);
|
||||||
if options_.relative_irf
|
if options_.relative_irf
|
||||||
|
@ -429,9 +429,9 @@ if ~options_.nograph && ~options_.no_graph.posterior
|
||||||
fprintf(fidTeX,'\\label{Fig:BayesianIRF:%s:%d}\n',deblank(tit(ii,:)),figunumber);
|
fprintf(fidTeX,'\\label{Fig:BayesianIRF:%s:%d}\n',deblank(tit(ii,:)),figunumber);
|
||||||
fprintf(fidTeX,'\\end{figure}\n');
|
fprintf(fidTeX,'\\end{figure}\n');
|
||||||
fprintf(fidTeX,' \n');
|
fprintf(fidTeX,' \n');
|
||||||
|
|
||||||
subplotnum = 0;
|
subplotnum = 0;
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
fprintf(fidTeX,'%% End of TeX file.\n');
|
fprintf(fidTeX,'%% End of TeX file.\n');
|
||||||
|
@ -443,11 +443,11 @@ if ~options_.nograph && ~options_.no_graph.posterior
|
||||||
|
|
||||||
% Comment for testing!
|
% Comment for testing!
|
||||||
if ~isoctave
|
if ~isoctave
|
||||||
if isnumeric(options_.parallel) || (M_.exo_nbr*ceil(size(varlist,1)/MaxNumberOfPlotPerFigure))<8,
|
if isnumeric(options_.parallel) || (M_.exo_nbr*ceil(size(varlist,1)/MaxNumberOfPlotPerFigure))<8
|
||||||
[fout] = PosteriorIRF_core2(localVars,1,M_.exo_nbr,0);
|
[fout] = PosteriorIRF_core2(localVars,1,M_.exo_nbr,0);
|
||||||
else
|
else
|
||||||
isRemoteOctave = 0;
|
isRemoteOctave = 0;
|
||||||
for indPC=1:length(options_.parallel),
|
for indPC=1:length(options_.parallel)
|
||||||
isRemoteOctave = isRemoteOctave + (findstr(options_.parallel(indPC).MatlabOctavePath, 'octave'));
|
isRemoteOctave = isRemoteOctave + (findstr(options_.parallel(indPC).MatlabOctavePath, 'octave'));
|
||||||
end
|
end
|
||||||
if isRemoteOctave
|
if isRemoteOctave
|
||||||
|
|
|
@ -23,7 +23,7 @@ function myoutput=PosteriorIRF_core1(myinputs,fpar,B,whoiam, ThisMatlab)
|
||||||
% SPECIAL REQUIREMENTS.
|
% SPECIAL REQUIREMENTS.
|
||||||
% None.
|
% None.
|
||||||
%
|
%
|
||||||
% Copyright (C) 2006-2016 Dynare Team
|
% Copyright (C) 2006-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -43,7 +43,7 @@ function myoutput=PosteriorIRF_core1(myinputs,fpar,B,whoiam, ThisMatlab)
|
||||||
|
|
||||||
global options_ estim_params_ oo_ M_ bayestopt_ dataset_ dataset_info
|
global options_ estim_params_ oo_ M_ bayestopt_ dataset_ dataset_info
|
||||||
|
|
||||||
if nargin<4,
|
if nargin<4
|
||||||
whoiam=0;
|
whoiam=0;
|
||||||
end
|
end
|
||||||
|
|
||||||
|
@ -55,7 +55,7 @@ irun =myinputs.irun;
|
||||||
irun2=myinputs.irun2;
|
irun2=myinputs.irun2;
|
||||||
npar=myinputs.npar;
|
npar=myinputs.npar;
|
||||||
type=myinputs.type;
|
type=myinputs.type;
|
||||||
if ~strcmpi(type,'prior'),
|
if ~strcmpi(type,'prior')
|
||||||
x=myinputs.x;
|
x=myinputs.x;
|
||||||
end
|
end
|
||||||
|
|
||||||
|
@ -102,7 +102,7 @@ end
|
||||||
RemoteFlag = 0;
|
RemoteFlag = 0;
|
||||||
|
|
||||||
if whoiam
|
if whoiam
|
||||||
if Parallel(ThisMatlab).Local==0,
|
if Parallel(ThisMatlab).Local==0
|
||||||
RemoteFlag =1;
|
RemoteFlag =1;
|
||||||
end
|
end
|
||||||
prct0={0,whoiam,Parallel(ThisMatlab)};
|
prct0={0,whoiam,Parallel(ThisMatlab)};
|
||||||
|
@ -165,7 +165,7 @@ while fpar<B
|
||||||
elseif info(1) == 5
|
elseif info(1) == 5
|
||||||
errordef = 'Rank condition is not satisfied';
|
errordef = 'Rank condition is not satisfied';
|
||||||
end
|
end
|
||||||
if strcmpi(type,'prior'),
|
if strcmpi(type,'prior')
|
||||||
disp(['PosteriorIRF :: Dynare is unable to solve the model (' errordef ')'])
|
disp(['PosteriorIRF :: Dynare is unable to solve the model (' errordef ')'])
|
||||||
continue
|
continue
|
||||||
else
|
else
|
||||||
|
@ -240,9 +240,9 @@ while fpar<B
|
||||||
else
|
else
|
||||||
stock_irf_bvardsge(:,:,:,IRUN) = reshape(tmp_dsgevar,options_.irf,dataset_.vobs,M_.exo_nbr);
|
stock_irf_bvardsge(:,:,:,IRUN) = reshape(tmp_dsgevar,options_.irf,dataset_.vobs,M_.exo_nbr);
|
||||||
instr = [MhDirectoryName '/' M_.fname '_irf_bvardsge' ...
|
instr = [MhDirectoryName '/' M_.fname '_irf_bvardsge' ...
|
||||||
int2str(NumberOfIRFfiles_dsgevar) '.mat stock_irf_bvardsge;'];,
|
int2str(NumberOfIRFfiles_dsgevar) '.mat stock_irf_bvardsge;'];
|
||||||
eval(['save ' instr]);
|
eval(['save ' instr]);
|
||||||
if RemoteFlag==1,
|
if RemoteFlag==1
|
||||||
OutputFileName_bvardsge = [OutputFileName_bvardsge; {[MhDirectoryName filesep], [M_.fname '_irf_bvardsge' int2str(NumberOfIRFfiles_dsgevar) '.mat']}];
|
OutputFileName_bvardsge = [OutputFileName_bvardsge; {[MhDirectoryName filesep], [M_.fname '_irf_bvardsge' int2str(NumberOfIRFfiles_dsgevar) '.mat']}];
|
||||||
end
|
end
|
||||||
NumberOfIRFfiles_dsgevar = NumberOfIRFfiles_dsgevar+1;
|
NumberOfIRFfiles_dsgevar = NumberOfIRFfiles_dsgevar+1;
|
||||||
|
@ -259,14 +259,14 @@ while fpar<B
|
||||||
int2str(NumberOfIRFfiles_dsgevar) '.mat stock_irf_bvardsge;'];
|
int2str(NumberOfIRFfiles_dsgevar) '.mat stock_irf_bvardsge;'];
|
||||||
eval(['save ' instr]);
|
eval(['save ' instr]);
|
||||||
NumberOfIRFfiles_dsgevar = NumberOfIRFfiles_dsgevar+1;
|
NumberOfIRFfiles_dsgevar = NumberOfIRFfiles_dsgevar+1;
|
||||||
if RemoteFlag==1,
|
if RemoteFlag==1
|
||||||
OutputFileName_bvardsge = [OutputFileName_bvardsge; {[MhDirectoryName filesep], [M_.fname '_irf_bvardsge' int2str(NumberOfIRFfiles_dsgevar) '.mat']}];
|
OutputFileName_bvardsge = [OutputFileName_bvardsge; {[MhDirectoryName filesep], [M_.fname '_irf_bvardsge' int2str(NumberOfIRFfiles_dsgevar) '.mat']}];
|
||||||
end
|
end
|
||||||
irun = 0;
|
irun = 0;
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
save([MhDirectoryName '/' M_.fname '_irf_dsge' int2str(NumberOfIRFfiles_dsge) '.mat'],'stock_irf_dsge');
|
save([MhDirectoryName '/' M_.fname '_irf_dsge' int2str(NumberOfIRFfiles_dsge) '.mat'],'stock_irf_dsge');
|
||||||
if RemoteFlag==1,
|
if RemoteFlag==1
|
||||||
OutputFileName_dsge = [OutputFileName_dsge; {[MhDirectoryName filesep], [M_.fname '_irf_dsge' int2str(NumberOfIRFfiles_dsge) '.mat']}];
|
OutputFileName_dsge = [OutputFileName_dsge; {[MhDirectoryName filesep], [M_.fname '_irf_dsge' int2str(NumberOfIRFfiles_dsge) '.mat']}];
|
||||||
end
|
end
|
||||||
NumberOfIRFfiles_dsge = NumberOfIRFfiles_dsge+1;
|
NumberOfIRFfiles_dsge = NumberOfIRFfiles_dsge+1;
|
||||||
|
@ -278,7 +278,7 @@ while fpar<B
|
||||||
end
|
end
|
||||||
stock = stock_param;
|
stock = stock_param;
|
||||||
save([MhDirectoryName '/' M_.fname '_param_irf' int2str(ifil2) '.mat'],'stock');
|
save([MhDirectoryName '/' M_.fname '_param_irf' int2str(ifil2) '.mat'],'stock');
|
||||||
if RemoteFlag==1,
|
if RemoteFlag==1
|
||||||
OutputFileName_param = [OutputFileName_param; {[MhDirectoryName filesep], [M_.fname '_param_irf' int2str(ifil2) '.mat']}];
|
OutputFileName_param = [OutputFileName_param; {[MhDirectoryName filesep], [M_.fname '_param_irf' int2str(ifil2) '.mat']}];
|
||||||
end
|
end
|
||||||
ifil2 = ifil2 + 1;
|
ifil2 = ifil2 + 1;
|
||||||
|
|
|
@ -2,10 +2,10 @@ function myoutput=PosteriorIRF_core2(myinputs,fpar,npar,whoiam,ThisMatlab)
|
||||||
% function myoutput=PosteriorIRF_core2(myinputs,fpar,npar,whoiam, ThisMatlab)
|
% function myoutput=PosteriorIRF_core2(myinputs,fpar,npar,whoiam, ThisMatlab)
|
||||||
% Generates the Posterior IRFs plot from the IRFs generated in
|
% Generates the Posterior IRFs plot from the IRFs generated in
|
||||||
% PosteriorIRF_core1
|
% PosteriorIRF_core1
|
||||||
%
|
%
|
||||||
% PARALLEL CONTEXT
|
% PARALLEL CONTEXT
|
||||||
% Performs in parallel execution a portion of the PosteriorIRF.m code.
|
% Performs in parallel execution a portion of the PosteriorIRF.m code.
|
||||||
% For more information, see the comment in posterior_sampler_core.m
|
% For more information, see the comment in posterior_sampler_core.m
|
||||||
% function.
|
% function.
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
|
@ -49,7 +49,7 @@ function myoutput=PosteriorIRF_core2(myinputs,fpar,npar,whoiam,ThisMatlab)
|
||||||
|
|
||||||
global options_ M_
|
global options_ M_
|
||||||
|
|
||||||
if nargin<4,
|
if nargin<4
|
||||||
whoiam=0;
|
whoiam=0;
|
||||||
end
|
end
|
||||||
|
|
||||||
|
@ -85,8 +85,8 @@ end
|
||||||
DirectoryName = CheckPath('Output',M_.dname);
|
DirectoryName = CheckPath('Output',M_.dname);
|
||||||
|
|
||||||
RemoteFlag = 0;
|
RemoteFlag = 0;
|
||||||
if whoiam,
|
if whoiam
|
||||||
if Parallel(ThisMatlab).Local==0,
|
if Parallel(ThisMatlab).Local==0
|
||||||
RemoteFlag =1;
|
RemoteFlag =1;
|
||||||
end
|
end
|
||||||
prct0={0,whoiam,Parallel(ThisMatlab)};
|
prct0={0,whoiam,Parallel(ThisMatlab)};
|
||||||
|
@ -96,7 +96,7 @@ end
|
||||||
OutputFileName={};
|
OutputFileName={};
|
||||||
|
|
||||||
subplotnum = 0;
|
subplotnum = 0;
|
||||||
for i=fpar:npar,
|
for i=fpar:npar
|
||||||
figunumber = 0;
|
figunumber = 0;
|
||||||
|
|
||||||
for j=1:nvar
|
for j=1:nvar
|
||||||
|
@ -153,13 +153,13 @@ for i=fpar:npar,
|
||||||
if subplotnum == MaxNumberOfPlotPerFigure || (j == nvar && subplotnum> 0)
|
if subplotnum == MaxNumberOfPlotPerFigure || (j == nvar && subplotnum> 0)
|
||||||
figunumber = figunumber+1;
|
figunumber = figunumber+1;
|
||||||
dyn_saveas(hh,[DirectoryName '/' M_.fname '_Bayesian_IRF_' deblank(tit(i,:)) '_' int2str(figunumber)],options_.nodisplay,options_.graph_format);
|
dyn_saveas(hh,[DirectoryName '/' M_.fname '_Bayesian_IRF_' deblank(tit(i,:)) '_' int2str(figunumber)],options_.nodisplay,options_.graph_format);
|
||||||
if RemoteFlag==1,
|
if RemoteFlag==1
|
||||||
OutputFileName = [OutputFileName; {[DirectoryName,filesep], [M_.fname '_Bayesian_IRF_' deblank(tit(i,:)) '_' int2str(figunumber) '.*']}];
|
OutputFileName = [OutputFileName; {[DirectoryName,filesep], [M_.fname '_Bayesian_IRF_' deblank(tit(i,:)) '_' int2str(figunumber) '.*']}];
|
||||||
end
|
end
|
||||||
subplotnum = 0;
|
subplotnum = 0;
|
||||||
end
|
end
|
||||||
end% loop over selected endo_var
|
end% loop over selected endo_var
|
||||||
if whoiam,
|
if whoiam
|
||||||
fprintf('Done! \n');
|
fprintf('Done! \n');
|
||||||
waitbarString = [ 'Exog. shocks ' int2str(i) '/' int2str(npar) ' done.'];
|
waitbarString = [ 'Exog. shocks ' int2str(i) '/' int2str(npar) ' done.'];
|
||||||
% fMessageStatus((i-fpar+1)/(npar-fpar+1),whoiam,waitbarString, waitbarTitle, Parallel(ThisMatlab));
|
% fMessageStatus((i-fpar+1)/(npar-fpar+1),whoiam,waitbarString, waitbarTitle, Parallel(ThisMatlab));
|
||||||
|
|
|
@ -18,14 +18,14 @@ function ReshapeMatFiles(type, type2)
|
||||||
% posterior
|
% posterior
|
||||||
% gsa
|
% gsa
|
||||||
% prior
|
% prior
|
||||||
%
|
%
|
||||||
% OUTPUTS:
|
% OUTPUTS:
|
||||||
% none
|
% none
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2003-2011 Dynare Team
|
% Copyright (C) 2003-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -44,15 +44,15 @@ function ReshapeMatFiles(type, type2)
|
||||||
|
|
||||||
global M_ options_
|
global M_ options_
|
||||||
|
|
||||||
if nargin==1,
|
if nargin==1
|
||||||
MhDirectoryName = [ CheckPath('metropolis',M_.dname) filesep ];
|
MhDirectoryName = [ CheckPath('metropolis',M_.dname) filesep ];
|
||||||
else
|
else
|
||||||
if strcmpi(type2,'posterior')
|
if strcmpi(type2,'posterior')
|
||||||
MhDirectoryName = [CheckPath('metropolis',M_.dname) filesep ];
|
MhDirectoryName = [CheckPath('metropolis',M_.dname) filesep ];
|
||||||
elseif strcmpi(type2,'gsa')
|
elseif strcmpi(type2,'gsa')
|
||||||
if options_.opt_gsa.morris==1,
|
if options_.opt_gsa.morris==1
|
||||||
MhDirectoryName = [CheckPath('gsa/screen',M_.dname) filesep ];
|
MhDirectoryName = [CheckPath('gsa/screen',M_.dname) filesep ];
|
||||||
elseif options_.opt_gsa.morris==2,
|
elseif options_.opt_gsa.morris==2
|
||||||
MhDirectoryName = [CheckPath('gsa/identif',M_.dname) filesep ];
|
MhDirectoryName = [CheckPath('gsa/identif',M_.dname) filesep ];
|
||||||
elseif options_.opt_gsa.pprior
|
elseif options_.opt_gsa.pprior
|
||||||
MhDirectoryName = [CheckPath(['gsa' filesep 'prior'],M_.dname) filesep ];
|
MhDirectoryName = [CheckPath(['gsa' filesep 'prior'],M_.dname) filesep ];
|
||||||
|
@ -61,17 +61,17 @@ else
|
||||||
end
|
end
|
||||||
else
|
else
|
||||||
MhDirectoryName = [CheckPath('prior',M_.dname) filesep ];
|
MhDirectoryName = [CheckPath('prior',M_.dname) filesep ];
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
switch type
|
switch type
|
||||||
case 'irf_dsge'
|
case 'irf_dsge'
|
||||||
CAPtype = 'IRF_DSGE';
|
CAPtype = 'IRF_DSGE';
|
||||||
TYPEsize = [ options_.irf , size(options_.varlist,1) , M_.exo_nbr ];
|
TYPEsize = [ options_.irf , size(options_.varlist,1) , M_.exo_nbr ];
|
||||||
TYPEarray = 4;
|
TYPEarray = 4;
|
||||||
case 'irf_bvardsge'
|
case 'irf_bvardsge'
|
||||||
CAPtype = 'IRF_BVARDSGE';
|
CAPtype = 'IRF_BVARDSGE';
|
||||||
TYPEsize = [ options_.irf , length(options_.varobs) , M_.exo_nbr ];
|
TYPEsize = [ options_.irf , length(options_.varobs) , M_.exo_nbr ];
|
||||||
TYPEarray = 4;
|
TYPEarray = 4;
|
||||||
case 'smooth'
|
case 'smooth'
|
||||||
CAPtype = 'SMOOTH';
|
CAPtype = 'SMOOTH';
|
||||||
TYPEsize = [ M_.endo_nbr , options_.nobs ];
|
TYPEsize = [ M_.endo_nbr , options_.nobs ];
|
||||||
|
@ -134,7 +134,7 @@ switch TYPEarray
|
||||||
eval(['idx = idx + size(stock_' type ',4);'])
|
eval(['idx = idx + size(stock_' type ',4);'])
|
||||||
end
|
end
|
||||||
%eval(['STOCK_' CAPtype ' = sort(STOCK_' CAPtype ',4);'])
|
%eval(['STOCK_' CAPtype ' = sort(STOCK_' CAPtype ',4);'])
|
||||||
save([MhDirectoryName M_.fname '_' CAPtype 's' int2str(NumberOfTYPEfiles-foffset+1) '.mat'],['STOCK_' CAPtype]);
|
save([MhDirectoryName M_.fname '_' CAPtype 's' int2str(NumberOfTYPEfiles-foffset+1) '.mat'],['STOCK_' CAPtype]);
|
||||||
end
|
end
|
||||||
else
|
else
|
||||||
load([MhDirectoryName M_.fname '_' type '1.mat']);
|
load([MhDirectoryName M_.fname '_' type '1.mat']);
|
||||||
|
@ -176,7 +176,7 @@ switch TYPEarray
|
||||||
load([MhDirectoryName M_.fname '_' type '1.mat']);
|
load([MhDirectoryName M_.fname '_' type '1.mat']);
|
||||||
%eval(['STOCK_' CAPtype ' = sort(stock_' type ',3);'])
|
%eval(['STOCK_' CAPtype ' = sort(stock_' type ',3);'])
|
||||||
eval(['STOCK_' CAPtype ' = stock_' type ';'])
|
eval(['STOCK_' CAPtype ' = stock_' type ';'])
|
||||||
save([MhDirectoryName M_.fname '_' CAPtype 's' int2str(1) '.mat'],['STOCK_' CAPtype ]);
|
save([MhDirectoryName M_.fname '_' CAPtype 's' int2str(1) '.mat'],['STOCK_' CAPtype ]);
|
||||||
end
|
end
|
||||||
% Original file format may be useful in some cases...
|
% Original file format may be useful in some cases...
|
||||||
% for file = 1:NumberOfTYPEfiles
|
% for file = 1:NumberOfTYPEfiles
|
||||||
|
|
|
@ -3,15 +3,15 @@ function [fval,info,exit_flag,DLIK,Hess,SteadyState,trend_coeff] = TaRB_optimiz
|
||||||
% Wrapper function for target function used in TaRB algorithm; reassembles
|
% Wrapper function for target function used in TaRB algorithm; reassembles
|
||||||
% full parameter vector before calling target function
|
% full parameter vector before calling target function
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% o optpar [double] (p_opt*1) vector of subset of parameters to be considered
|
% o optpar [double] (p_opt*1) vector of subset of parameters to be considered
|
||||||
% o par_vector [double] (p*1) full vector of parameters
|
% o par_vector [double] (p*1) full vector of parameters
|
||||||
% o parameterindices [double] (p_opt*1) index of optpar entries in
|
% o parameterindices [double] (p_opt*1) index of optpar entries in
|
||||||
% par_vector
|
% par_vector
|
||||||
% o TargetFun [char] string specifying the name of the objective
|
% o TargetFun [char] string specifying the name of the objective
|
||||||
% function (posterior kernel).
|
% function (posterior kernel).
|
||||||
% o varargin [structure] other inputs of target function
|
% o varargin [structure] other inputs of target function
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% o fval [scalar] value of (minus) the likelihood.
|
% o fval [scalar] value of (minus) the likelihood.
|
||||||
% o info [double] (p*2) error code vector
|
% o info [double] (p*2) error code vector
|
||||||
|
@ -20,8 +20,9 @@ function [fval,info,exit_flag,DLIK,Hess,SteadyState,trend_coeff] = TaRB_optimiz
|
||||||
% o Hess [double] (p*p) asymptotic Hessian matrix.
|
% o Hess [double] (p*p) asymptotic Hessian matrix.
|
||||||
% o SteadyState [double] Vector of doubles, steady state level for the endogenous variables.
|
% o SteadyState [double] Vector of doubles, steady state level for the endogenous variables.
|
||||||
% o trend_coeff [double] Matrix of doubles, coefficients of the deterministic trend in the measurement equation
|
% o trend_coeff [double] Matrix of doubles, coefficients of the deterministic trend in the measurement equation
|
||||||
%
|
%
|
||||||
% Copyright (C) 2015-16 Dynare Team
|
|
||||||
|
% Copyright (C) 2015-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -40,4 +41,3 @@ function [fval,info,exit_flag,DLIK,Hess,SteadyState,trend_coeff] = TaRB_optimiz
|
||||||
|
|
||||||
par_vector(parameterindices,:)=optpar; %reassemble parameter
|
par_vector(parameterindices,:)=optpar; %reassemble parameter
|
||||||
[fval,info,exit_flag,DLIK,Hess,SteadyState,trend_coeff] = feval(TargetFun,par_vector,varargin{:}); %call target function
|
[fval,info,exit_flag,DLIK,Hess,SteadyState,trend_coeff] = feval(TargetFun,par_vector,varargin{:}); %call target function
|
||||||
|
|
||||||
|
|
|
@ -2,19 +2,19 @@ function [] = Tracing()
|
||||||
% DESCRIPTION
|
% DESCRIPTION
|
||||||
% This function is used to test the correct execution of a matlab section
|
% This function is used to test the correct execution of a matlab section
|
||||||
% on remote machine.
|
% on remote machine.
|
||||||
%
|
%
|
||||||
% If no error happen the function simply create a file.
|
% If no error happen the function simply create a file.
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% ...
|
% ...
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% ...
|
% ...
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2010 Dynare Team
|
% Copyright (C) 2010-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
|
|
@ -1,15 +1,15 @@
|
||||||
function [oo_] = UnivariateSpectralDensity(M_,oo_,options_,var_list)
|
function [oo_] = UnivariateSpectralDensity(M_,oo_,options_,var_list)
|
||||||
% This function computes the theoretical spectral density of each
|
% This function computes the theoretical spectral density of each
|
||||||
% endogenous variable declared in var_list. Results are stored in
|
% endogenous variable declared in var_list. Results are stored in
|
||||||
% oo_.SpectralDensity and may be plotted. Plots are saved into the
|
% oo_.SpectralDensity and may be plotted. Plots are saved into the
|
||||||
% graphs-folder.
|
% graphs-folder.
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% M_ [structure] Dynare's model structure
|
% M_ [structure] Dynare's model structure
|
||||||
% oo_ [structure] Dynare's results structure
|
% oo_ [structure] Dynare's results structure
|
||||||
% options_ [structure] Dynare's options structure
|
% options_ [structure] Dynare's options structure
|
||||||
% var_list [integer] Vector of indices for a subset of variables.
|
% var_list [integer] Vector of indices for a subset of variables.
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% oo_ [structure] Dynare's results structure,
|
% oo_ [structure] Dynare's results structure,
|
||||||
% containing the subfield
|
% containing the subfield
|
||||||
|
@ -17,7 +17,7 @@ function [oo_] = UnivariateSpectralDensity(M_,oo_,options_,var_list)
|
||||||
% and density, which are of size nvar*ngrid.
|
% and density, which are of size nvar*ngrid.
|
||||||
%
|
%
|
||||||
|
|
||||||
% Adapted from th_autocovariances.m.
|
% Adapted from th_autocovariances.m.
|
||||||
|
|
||||||
% Copyright (C) 2006-2017 Dynare Team
|
% Copyright (C) 2006-2017 Dynare Team
|
||||||
%
|
%
|
||||||
|
@ -38,7 +38,7 @@ function [oo_] = UnivariateSpectralDensity(M_,oo_,options_,var_list)
|
||||||
|
|
||||||
|
|
||||||
if options_.order > 1
|
if options_.order > 1
|
||||||
disp('UnivariateSpectralDensity :: I Cannot compute the theoretical spectral density')
|
disp('UnivariateSpectralDensity :: I Cannot compute the theoretical spectral density')
|
||||||
disp('with a second order approximation of the DSGE model!')
|
disp('with a second order approximation of the DSGE model!')
|
||||||
disp('Please set order = 1. I abort')
|
disp('Please set order = 1. I abort')
|
||||||
return
|
return
|
||||||
|
@ -102,7 +102,7 @@ if ~isempty(u)
|
||||||
ivar = oo_.dr.order_var(iky);
|
ivar = oo_.dr.order_var(iky);
|
||||||
end
|
end
|
||||||
|
|
||||||
iky = iv(ivar);
|
iky = iv(ivar);
|
||||||
aa = ghx(iky,:);
|
aa = ghx(iky,:);
|
||||||
bb = ghu(iky,:);
|
bb = ghu(iky,:);
|
||||||
ngrid = options_.hp_ngrid; %number of grid points
|
ngrid = options_.hp_ngrid; %number of grid points
|
||||||
|
@ -122,7 +122,7 @@ elseif ~(options_.hp_filter == 0 && ~options_.bandpass.indicator) && options_.ba
|
||||||
filter_gain(freqs<=-2*pi/lowest_periodicity+2*pi & freqs>=-2*pi/highest_periodicity+2*pi)=1;
|
filter_gain(freqs<=-2*pi/lowest_periodicity+2*pi & freqs>=-2*pi/highest_periodicity+2*pi)=1;
|
||||||
elseif ~(options_.hp_filter == 0 && ~options_.bandpass.indicator) && ~options_.bandpass.indicator %filter with HP-filter
|
elseif ~(options_.hp_filter == 0 && ~options_.bandpass.indicator) && ~options_.bandpass.indicator %filter with HP-filter
|
||||||
lambda = options_.hp_filter;
|
lambda = options_.hp_filter;
|
||||||
filter_gain = 4*lambda*(1 - cos(freqs)).^2 ./ (1 + 4*lambda*(1 - cos(freqs)).^2);
|
filter_gain = 4*lambda*(1 - cos(freqs)).^2 ./ (1 + 4*lambda*(1 - cos(freqs)).^2);
|
||||||
end
|
end
|
||||||
|
|
||||||
mathp_col = NaN(ngrid,length(ivar)^2);
|
mathp_col = NaN(ngrid,length(ivar)^2);
|
||||||
|
@ -134,12 +134,12 @@ for ig = 1:ngrid
|
||||||
g_omega = [aa*tneg(ig) bb]*f_omega*[aa'*tpos(ig); bb']; % selected variables
|
g_omega = [aa*tneg(ig) bb]*f_omega*[aa'*tpos(ig); bb']; % selected variables
|
||||||
f_hp = filter_gain(ig)^2*g_omega; % spectral density of selected filtered series
|
f_hp = filter_gain(ig)^2*g_omega; % spectral density of selected filtered series
|
||||||
mathp_col(ig,:) = (f_hp(:))'; % store as matrix row
|
mathp_col(ig,:) = (f_hp(:))'; % store as matrix row
|
||||||
end;
|
end
|
||||||
|
|
||||||
f = zeros(nvar,ngrid);
|
f = zeros(nvar,ngrid);
|
||||||
for i=1:nvar
|
for i=1:nvar
|
||||||
f(i,:) = real(mathp_col(:,(i-1)*nvar+i)); %read out spectral density
|
f(i,:) = real(mathp_col(:,(i-1)*nvar+i)); %read out spectral density
|
||||||
end
|
end
|
||||||
|
|
||||||
oo_.SpectralDensity.freqs=freqs;
|
oo_.SpectralDensity.freqs=freqs;
|
||||||
oo_.SpectralDensity.density=f;
|
oo_.SpectralDensity.density=f;
|
||||||
|
@ -164,7 +164,7 @@ if options_.nograph == 0
|
||||||
xlabel('0 \leq \omega \leq \pi')
|
xlabel('0 \leq \omega \leq \pi')
|
||||||
ylabel('f(\omega)')
|
ylabel('f(\omega)')
|
||||||
box on
|
box on
|
||||||
axis tight
|
axis tight
|
||||||
dyn_saveas(hh,[M_.fname ,filesep,'graphs', filesep, 'SpectralDensity_' deblank(M_.endo_names(ivar(i),:))],options_.nodisplay,options_.graph_format)
|
dyn_saveas(hh,[M_.fname ,filesep,'graphs', filesep, 'SpectralDensity_' deblank(M_.endo_names(ivar(i),:))],options_.nodisplay,options_.graph_format)
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
|
@ -1,7 +1,7 @@
|
||||||
function WriteShockDecomp2Excel(z,shock_names,endo_names,i_var,initial_date,DynareModel,DynareOptions,opts_decomp)
|
function WriteShockDecomp2Excel(z,shock_names,endo_names,i_var,initial_date,DynareModel,DynareOptions,opts_decomp)
|
||||||
%function WriteShockDecomp2Excel(z,shock_names,endo_names,i_var,initial_date,DynareModel,DynareOptions)
|
%function WriteShockDecomp2Excel(z,shock_names,endo_names,i_var,initial_date,DynareModel,DynareOptions)
|
||||||
% Saves the results from the shock_decomposition command to xls
|
% Saves the results from the shock_decomposition command to xls
|
||||||
%
|
%
|
||||||
% Inputs
|
% Inputs
|
||||||
% z [n_var*(nshock+2)*nperiods] shock decomposition array, see shock_decomposition.m for details
|
% z [n_var*(nshock+2)*nperiods] shock decomposition array, see shock_decomposition.m for details
|
||||||
% shock_names [endo_nbr*string length] shock names from M_.exo_names
|
% shock_names [endo_nbr*string length] shock names from M_.exo_names
|
||||||
|
@ -42,14 +42,14 @@ end
|
||||||
% number of components equals number of shocks + 1 (initial conditions)
|
% number of components equals number of shocks + 1 (initial conditions)
|
||||||
comp_nbr = size(z,2)-1;
|
comp_nbr = size(z,2)-1;
|
||||||
|
|
||||||
if nargin==8 ,
|
if nargin==8
|
||||||
if isfield(opts_decomp,'steady_state')
|
if isfield(opts_decomp,'steady_state')
|
||||||
SteadyState = opts_decomp.steady_state;
|
SteadyState = opts_decomp.steady_state;
|
||||||
end
|
end
|
||||||
if isfield(opts_decomp,'fig_mode') && ~isempty(opts_decomp.fig_mode)
|
if isfield(opts_decomp,'fig_mode') && ~isempty(opts_decomp.fig_mode)
|
||||||
fig_mode = opts_decomp.fig_mode;
|
fig_mode = opts_decomp.fig_mode;
|
||||||
fig_mode1 = ['_' fig_mode];
|
fig_mode1 = ['_' fig_mode];
|
||||||
fig_mode = [fig_mode '_'];
|
fig_mode = [fig_mode '_'];
|
||||||
end
|
end
|
||||||
if isfield(opts_decomp,'screen_shocks')
|
if isfield(opts_decomp,'screen_shocks')
|
||||||
if use_shock_groups
|
if use_shock_groups
|
||||||
|
@ -61,14 +61,14 @@ if nargin==8 ,
|
||||||
if isfield(opts_decomp,'fig_name')
|
if isfield(opts_decomp,'fig_name')
|
||||||
fig_name = opts_decomp.fig_name;
|
fig_name = opts_decomp.fig_name;
|
||||||
% fig_name = ['_' fig_name];
|
% fig_name = ['_' fig_name];
|
||||||
fig_name1 = [fig_name];
|
fig_name1 = [fig_name];
|
||||||
fig_name = [fig_name '_'];
|
fig_name = [fig_name '_'];
|
||||||
end
|
end
|
||||||
if screen_shocks
|
if screen_shocks
|
||||||
fig_name1 = [fig_name1 '_screen'];
|
fig_name1 = [fig_name1 '_screen'];
|
||||||
fig_name = [fig_name 'screen_'];
|
fig_name = [fig_name 'screen_'];
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
||||||
gend = size(z,3);
|
gend = size(z,3);
|
||||||
|
@ -84,7 +84,7 @@ end
|
||||||
nvar = length(i_var);
|
nvar = length(i_var);
|
||||||
|
|
||||||
labels = char(char(shock_names),'Initial values');
|
labels = char(char(shock_names),'Initial values');
|
||||||
if ~(screen_shocks && comp_nbr>18),
|
if ~(screen_shocks && comp_nbr>18)
|
||||||
screen_shocks=0;
|
screen_shocks=0;
|
||||||
end
|
end
|
||||||
comp_nbr0=comp_nbr;
|
comp_nbr0=comp_nbr;
|
||||||
|
@ -92,14 +92,14 @@ comp_nbr0=comp_nbr;
|
||||||
for j=1:nvar
|
for j=1:nvar
|
||||||
d0={};
|
d0={};
|
||||||
z1 = squeeze(z(i_var(j),:,:));
|
z1 = squeeze(z(i_var(j),:,:));
|
||||||
if screen_shocks,
|
if screen_shocks
|
||||||
[junk, isort] = sort(mean(abs(z1(1:end-2,:)')), 'descend');
|
[junk, isort] = sort(mean(abs(z1(1:end-2,:)')), 'descend');
|
||||||
labels = char(char(shock_names(isort(1:16),:)),'Others', 'Initial values');
|
labels = char(char(shock_names(isort(1:16),:)),'Others', 'Initial values');
|
||||||
zres = sum(z1(isort(17:end),:),1);
|
zres = sum(z1(isort(17:end),:),1);
|
||||||
z1 = [z1(isort(1:16),:); zres; z1(comp_nbr0:end,:)];
|
z1 = [z1(isort(1:16),:); zres; z1(comp_nbr0:end,:)];
|
||||||
comp_nbr=18;
|
comp_nbr=18;
|
||||||
end
|
end
|
||||||
|
|
||||||
d0(1,:)=[{'Decomposition'} cellstr(labels(1:comp_nbr,:))' {'Smoot Var'}];
|
d0(1,:)=[{'Decomposition'} cellstr(labels(1:comp_nbr,:))' {'Smoot Var'}];
|
||||||
d0=[d0; num2cell([x' z1'])];
|
d0=[d0; num2cell([x' z1'])];
|
||||||
LastRow=size(d0,1);
|
LastRow=size(d0,1);
|
||||||
|
@ -107,12 +107,12 @@ for j=1:nvar
|
||||||
d0(LastRow+2,1)={'Legend.'};
|
d0(LastRow+2,1)={'Legend.'};
|
||||||
d0(LastRow+2,2)={'Shocks include:'};
|
d0(LastRow+2,2)={'Shocks include:'};
|
||||||
d0(LastRow+3:LastRow+3+comp_nbr-1,1)=cellstr(labels(1:comp_nbr,:));
|
d0(LastRow+3:LastRow+3+comp_nbr-1,1)=cellstr(labels(1:comp_nbr,:));
|
||||||
for ic=1:comp_nbr,
|
for ic=1:comp_nbr
|
||||||
group_members = shock_groups.(shock_ind{ic}).shocks;
|
group_members = shock_groups.(shock_ind{ic}).shocks;
|
||||||
d0(LastRow+2+ic,2:1+length(group_members))=group_members;
|
d0(LastRow+2+ic,2:1+length(group_members))=group_members;
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
warning off
|
warning off
|
||||||
if ~ismac
|
if ~ismac
|
||||||
[STATUS,MESSAGE] = xlswrite([DynareModel.fname,'_shock_decomposition',fig_mode,fig_name1],d0,deblank(endo_names(i_var(j),:)));
|
[STATUS,MESSAGE] = xlswrite([DynareModel.fname,'_shock_decomposition',fig_mode,fig_name1],d0,deblank(endo_names(i_var(j),:)));
|
||||||
|
@ -122,6 +122,5 @@ for j=1:nvar
|
||||||
warning on
|
warning on
|
||||||
|
|
||||||
clear d0
|
clear d0
|
||||||
|
|
||||||
end
|
|
||||||
|
|
||||||
|
end
|
||||||
|
|
|
@ -1,5 +1,22 @@
|
||||||
function title=add_filter_subtitle(title,options_)
|
function title=add_filter_subtitle(title,options_)
|
||||||
|
|
||||||
|
% Copyright (C) 2015-2017 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/>.
|
||||||
|
|
||||||
if ~options_.hp_filter && ~options_.one_sided_hp_filter && ~options_.bandpass.indicator %do not filter
|
if ~options_.hp_filter && ~options_.one_sided_hp_filter && ~options_.bandpass.indicator %do not filter
|
||||||
%nothing to add here
|
%nothing to add here
|
||||||
elseif ~options_.hp_filter && ~options_.one_sided_hp_filter && options_.bandpass.indicator
|
elseif ~options_.hp_filter && ~options_.one_sided_hp_filter && options_.bandpass.indicator
|
||||||
|
@ -10,6 +27,6 @@ elseif options_.hp_filter && ~options_.one_sided_hp_filter && ~options_.bandpas
|
||||||
num2str(options_.hp_filter) ')'];
|
num2str(options_.hp_filter) ')'];
|
||||||
elseif ~options_.hp_filter && options_.one_sided_hp_filter && ~options_.bandpass.indicator
|
elseif ~options_.hp_filter && options_.one_sided_hp_filter && ~options_.bandpass.indicator
|
||||||
title = [title ' (One-sided HP filter, lambda = ' ...
|
title = [title ' (One-sided HP filter, lambda = ' ...
|
||||||
num2str(options_.one_sided_hp_filter) ')'];
|
num2str(options_.one_sided_hp_filter) ')'];
|
||||||
end
|
end
|
||||||
end
|
end
|
|
@ -1,28 +1,28 @@
|
||||||
function [z, endo_names, endo_names_tex, steady_state, i_var, oo_] = annualized_shock_decomposition(oo_, M_, options_, i_var, t0, t1, realtime_, vintage_, steady_state, q2a, cumfix)
|
function [z, endo_names, endo_names_tex, steady_state, i_var, oo_] = annualized_shock_decomposition(oo_, M_, options_, i_var, t0, t1, realtime_, vintage_, steady_state, q2a, cumfix)
|
||||||
% function oo_ = annualized_shock_decomposition(oo_,t0,options_.nobs);
|
% function oo_ = annualized_shock_decomposition(oo_,t0,options_.nobs);
|
||||||
% Computes annualized shocks contribution to a simulated trajectory. The fields set are
|
% Computes annualized shocks contribution to a simulated trajectory. The fields set are
|
||||||
% oo_.annualized_shock_decomposition, oo_.annualized_realtime_shock_decomposition,
|
% oo_.annualized_shock_decomposition, oo_.annualized_realtime_shock_decomposition,
|
||||||
% oo_.annualized_realtime_conditional_shock_decomposition and oo_.annualized_realtime_forecast_shock_decomposition.
|
% oo_.annualized_realtime_conditional_shock_decomposition and oo_.annualized_realtime_forecast_shock_decomposition.
|
||||||
% Subfields are arrays n_var by nshock+2 by nperiods. The
|
% Subfields are arrays n_var by nshock+2 by nperiods. The
|
||||||
% first nshock columns store the respective shock contributions, column n+1
|
% first nshock columns store the respective shock contributions, column n+1
|
||||||
% stores the role of the initial conditions, while column n+2 stores the
|
% stores the role of the initial conditions, while column n+2 stores the
|
||||||
% value of the smoothed variables. Both the variables and shocks are stored
|
% value of the smoothed variables. Both the variables and shocks are stored
|
||||||
% in the order of endo_names and M_.exo_names, respectively.
|
% in the order of endo_names and M_.exo_names, respectively.
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% oo_: [structure] Storage of results
|
% oo_: [structure] Storage of results
|
||||||
% M_: [structure] Storage of model
|
% M_: [structure] Storage of model
|
||||||
% opts: [structure] options for shock decomp
|
% opts: [structure] options for shock decomp
|
||||||
% i_var: [array] index of vars
|
% i_var: [array] index of vars
|
||||||
% t0: [integer] first period
|
% t0: [integer] first period
|
||||||
% t1: [integer] last period
|
% t1: [integer] last period
|
||||||
% realtime_: [integer]
|
% realtime_: [integer]
|
||||||
% vintage_: [integer]
|
% vintage_: [integer]
|
||||||
% steady_state: [array] steady state value of quarterly (log-) level vars
|
% steady_state: [array] steady state value of quarterly (log-) level vars
|
||||||
% q2a: [structure] info on q2a
|
% q2a: [structure] info on q2a
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% z: [matrix] shock decomp to plot
|
% z: [matrix] shock decomp to plot
|
||||||
% endo_names: [char] updated var names
|
% endo_names: [char] updated var names
|
||||||
% endo_names_tex: [char] updated TeX var names
|
% endo_names_tex: [char] updated TeX var names
|
||||||
% steady_state: [array] updated stady state of vars
|
% steady_state: [array] updated stady state of vars
|
||||||
|
@ -57,7 +57,7 @@ islog = q2a.islog;
|
||||||
aux = q2a.aux;
|
aux = q2a.aux;
|
||||||
aux0 = aux;
|
aux0 = aux;
|
||||||
cumfix = q2a.cumfix;
|
cumfix = q2a.cumfix;
|
||||||
% usual shock decomp
|
% usual shock decomp
|
||||||
if isstruct(oo_)
|
if isstruct(oo_)
|
||||||
% z = oo_.shock_decomposition;
|
% z = oo_.shock_decomposition;
|
||||||
myopts=options_;
|
myopts=options_;
|
||||||
|
@ -75,7 +75,7 @@ if isfield(q2a,'name')
|
||||||
if isfield(q2a,'tex_name')
|
if isfield(q2a,'tex_name')
|
||||||
mytex = q2a.tex_name;
|
mytex = q2a.tex_name;
|
||||||
end
|
end
|
||||||
if mytype==2,
|
if mytype==2
|
||||||
gtxt = ['PHI' mytxt]; % inflation rate
|
gtxt = ['PHI' mytxt]; % inflation rate
|
||||||
gtex = ['{\pi(' mytex ')}'];
|
gtex = ['{\pi(' mytex ')}'];
|
||||||
elseif mytype
|
elseif mytype
|
||||||
|
@ -101,7 +101,7 @@ if isstruct(aux)
|
||||||
end
|
end
|
||||||
yaux=aux.y;
|
yaux=aux.y;
|
||||||
end
|
end
|
||||||
if mytype==2,
|
if mytype==2
|
||||||
gtxt = 'PHI'; % inflation rate
|
gtxt = 'PHI'; % inflation rate
|
||||||
gtex = '\pi';
|
gtex = '\pi';
|
||||||
elseif mytype
|
elseif mytype
|
||||||
|
@ -115,7 +115,7 @@ nterms = size(z,2);
|
||||||
nfrcst = opts.forecast/4;
|
nfrcst = opts.forecast/4;
|
||||||
|
|
||||||
for j=1:nvar
|
for j=1:nvar
|
||||||
if j>1,
|
if j>1
|
||||||
endo_names = char(endo_names,[deblank(M_.endo_names(i_var(j),:)) '_A']);
|
endo_names = char(endo_names,[deblank(M_.endo_names(i_var(j),:)) '_A']);
|
||||||
endo_names_tex = char(endo_names_tex,['{' deblank(M_.endo_names_tex(i_var(j),:)) '}^A']);
|
endo_names_tex = char(endo_names_tex,['{' deblank(M_.endo_names_tex(i_var(j),:)) '}^A']);
|
||||||
gendo_names = char(gendo_names,[gtxt endo_names(j,:)]);
|
gendo_names = char(gendo_names,[gtxt endo_names(j,:)]);
|
||||||
|
@ -133,15 +133,15 @@ for j=1:nvar
|
||||||
gendo_names_tex = [gtex '(' deblank(endo_names_tex(j,:)) ')'];
|
gendo_names_tex = [gtex '(' deblank(endo_names_tex(j,:)) ')'];
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
for k =1:nterms,
|
for k =1:nterms
|
||||||
if isstruct(aux),
|
if isstruct(aux)
|
||||||
aux.y = squeeze(yaux(j,k,min((t0-3):-4:1):end));
|
aux.y = squeeze(yaux(j,k,min((t0-3):-4:1):end));
|
||||||
end
|
end
|
||||||
[za(j,k,:), steady_state_a(j,1), gza(j,k,:), steady_state_ga(j,1)] = ...
|
[za(j,k,:), steady_state_a(j,1), gza(j,k,:), steady_state_ga(j,1)] = ...
|
||||||
quarterly2annual(squeeze(z(j,k,min((t0-3):-4:1):end)),steady_state(j),GYTREND0,var_type,islog,aux);
|
quarterly2annual(squeeze(z(j,k,min((t0-3):-4:1):end)),steady_state(j),GYTREND0,var_type,islog,aux);
|
||||||
end
|
end
|
||||||
ztmp=squeeze(za(j,:,:));
|
ztmp=squeeze(za(j,:,:));
|
||||||
if cumfix==0,
|
if cumfix==0
|
||||||
zscale = sum(ztmp(1:end-1,:))./ztmp(end,:);
|
zscale = sum(ztmp(1:end-1,:))./ztmp(end,:);
|
||||||
ztmp(1:end-1,:) = ztmp(1:end-1,:)./repmat(zscale,[nterms-1,1]);
|
ztmp(1:end-1,:) = ztmp(1:end-1,:)./repmat(zscale,[nterms-1,1]);
|
||||||
else
|
else
|
||||||
|
@ -149,7 +149,7 @@ for j=1:nvar
|
||||||
ztmp(end-1,:) = ztmp(end-1,:) + zres;
|
ztmp(end-1,:) = ztmp(end-1,:) + zres;
|
||||||
end
|
end
|
||||||
gztmp=squeeze(gza(j,:,:));
|
gztmp=squeeze(gza(j,:,:));
|
||||||
if cumfix==0,
|
if cumfix==0
|
||||||
gscale = sum(gztmp(1:end-1,:))./ gztmp(end,:);
|
gscale = sum(gztmp(1:end-1,:))./ gztmp(end,:);
|
||||||
gztmp(1:end-1,:) = gztmp(1:end-1,:)./repmat(gscale,[nterms-1,1]);
|
gztmp(1:end-1,:) = gztmp(1:end-1,:)./repmat(gscale,[nterms-1,1]);
|
||||||
else
|
else
|
||||||
|
@ -160,7 +160,7 @@ for j=1:nvar
|
||||||
gza(j,:,:) = gztmp;
|
gza(j,:,:) = gztmp;
|
||||||
end
|
end
|
||||||
|
|
||||||
if q2a.plot ==1,
|
if q2a.plot ==1
|
||||||
z=gza;
|
z=gza;
|
||||||
endo_names = gendo_names;
|
endo_names = gendo_names;
|
||||||
endo_names_tex = gendo_names_tex;
|
endo_names_tex = gendo_names_tex;
|
||||||
|
@ -176,9 +176,9 @@ end
|
||||||
% end
|
% end
|
||||||
|
|
||||||
% realtime
|
% realtime
|
||||||
if realtime_ && isstruct(oo_) && isfield(oo_, 'realtime_shock_decomposition'),
|
if realtime_ && isstruct(oo_) && isfield(oo_, 'realtime_shock_decomposition')
|
||||||
init=1;
|
init=1;
|
||||||
for i=t0:4:t1,
|
for i=t0:4:t1
|
||||||
yr=floor(i/4);
|
yr=floor(i/4);
|
||||||
za=[];
|
za=[];
|
||||||
gza=[];
|
gza=[];
|
||||||
|
@ -197,18 +197,18 @@ if realtime_ && isstruct(oo_) && isfield(oo_, 'realtime_shock_decomposition'),
|
||||||
yaux=aux.y;
|
yaux=aux.y;
|
||||||
end
|
end
|
||||||
nterms = size(z,2);
|
nterms = size(z,2);
|
||||||
|
|
||||||
% z = oo_.realtime_shock_decomposition.(['time_' int2str(i)]);
|
% z = oo_.realtime_shock_decomposition.(['time_' int2str(i)]);
|
||||||
% z = z(i_var,:,:);
|
% z = z(i_var,:,:);
|
||||||
|
|
||||||
for j=1:nvar
|
for j=1:nvar
|
||||||
for k =nterms:-1:1,
|
for k =nterms:-1:1
|
||||||
% if k<nterms
|
% if k<nterms
|
||||||
% ztmp = squeeze(sum(z(j,[1:k-1,k+1:end-1],t0-4:end)));
|
% ztmp = squeeze(sum(z(j,[1:k-1,k+1:end-1],t0-4:end)));
|
||||||
% else
|
% else
|
||||||
ztmp = squeeze(z(j,k,min((t0-3):-4:1):end));
|
ztmp = squeeze(z(j,k,min((t0-3):-4:1):end));
|
||||||
% end
|
% end
|
||||||
if isstruct(aux),
|
if isstruct(aux)
|
||||||
aux.y = squeeze(yaux(j,k,min((t0-3):-4:1):end));
|
aux.y = squeeze(yaux(j,k,min((t0-3):-4:1):end));
|
||||||
end
|
end
|
||||||
[za(j,k,:), steady_state_a(j,1), gza(j,k,:), steady_state_ga(j,1)] = ...
|
[za(j,k,:), steady_state_a(j,1), gza(j,k,:), steady_state_ga(j,1)] = ...
|
||||||
|
@ -217,47 +217,47 @@ if realtime_ && isstruct(oo_) && isfield(oo_, 'realtime_shock_decomposition'),
|
||||||
% za(j,k,:) = za(j,end,:) - za(j,k,:);
|
% za(j,k,:) = za(j,end,:) - za(j,k,:);
|
||||||
% gza(j,k,:) = gza(j,end,:) - gza(j,k,:);
|
% gza(j,k,:) = gza(j,end,:) - gza(j,k,:);
|
||||||
% end
|
% end
|
||||||
|
|
||||||
end
|
end
|
||||||
|
|
||||||
ztmp=squeeze(za(j,:,:));
|
ztmp=squeeze(za(j,:,:));
|
||||||
|
|
||||||
if cumfix==0,
|
if cumfix==0
|
||||||
zscale = sum(ztmp(1:end-1,:))./ztmp(end,:);
|
zscale = sum(ztmp(1:end-1,:))./ztmp(end,:);
|
||||||
ztmp(1:end-1,:) = ztmp(1:end-1,:)./repmat(zscale,[nterms-1,1]);
|
ztmp(1:end-1,:) = ztmp(1:end-1,:)./repmat(zscale,[nterms-1,1]);
|
||||||
else
|
else
|
||||||
zres = ztmp(end,:)-sum(ztmp(1:end-1,:));
|
zres = ztmp(end,:)-sum(ztmp(1:end-1,:));
|
||||||
ztmp(end-1,:) = ztmp(end-1,:) + zres;
|
ztmp(end-1,:) = ztmp(end-1,:) + zres;
|
||||||
end
|
end
|
||||||
|
|
||||||
gztmp=squeeze(gza(j,:,:));
|
gztmp=squeeze(gza(j,:,:));
|
||||||
if cumfix==0,
|
if cumfix==0
|
||||||
gscale = sum(gztmp(1:end-1,:))./ gztmp(end,:);
|
gscale = sum(gztmp(1:end-1,:))./ gztmp(end,:);
|
||||||
gztmp(1:end-1,:) = gztmp(1:end-1,:)./repmat(gscale,[nterms-1,1]);
|
gztmp(1:end-1,:) = gztmp(1:end-1,:)./repmat(gscale,[nterms-1,1]);
|
||||||
else
|
else
|
||||||
gres = gztmp(end,:) - sum(gztmp(1:end-1,:));
|
gres = gztmp(end,:) - sum(gztmp(1:end-1,:));
|
||||||
gztmp(end-1,:) = gztmp(end-1,:)+gres;
|
gztmp(end-1,:) = gztmp(end-1,:)+gres;
|
||||||
end
|
end
|
||||||
|
|
||||||
za(j,:,:) = ztmp;
|
za(j,:,:) = ztmp;
|
||||||
gza(j,:,:) = gztmp;
|
gza(j,:,:) = gztmp;
|
||||||
end
|
end
|
||||||
|
|
||||||
if q2a.plot ==1,
|
if q2a.plot ==1
|
||||||
z=gza;
|
z=gza;
|
||||||
elseif q2a.plot == 2
|
elseif q2a.plot == 2
|
||||||
z=za;
|
z=za;
|
||||||
else
|
else
|
||||||
z=cat(1,za,gza);
|
z=cat(1,za,gza);
|
||||||
end
|
end
|
||||||
|
|
||||||
if init==1,
|
if init==1
|
||||||
oo_.annualized_realtime_shock_decomposition.pool = z;
|
oo_.annualized_realtime_shock_decomposition.pool = z;
|
||||||
else
|
else
|
||||||
oo_.annualized_realtime_shock_decomposition.pool(:,:,yr) = z(:,:,end-nfrcst);
|
oo_.annualized_realtime_shock_decomposition.pool(:,:,yr) = z(:,:,end-nfrcst);
|
||||||
end
|
end
|
||||||
oo_.annualized_realtime_shock_decomposition.(['yr_' int2str(yr)]) = z;
|
oo_.annualized_realtime_shock_decomposition.(['yr_' int2str(yr)]) = z;
|
||||||
|
|
||||||
if opts.forecast
|
if opts.forecast
|
||||||
oo_.annualized_realtime_forecast_shock_decomposition.(['yr_' int2str(yr)]) = z(:,:,end-nfrcst:end);
|
oo_.annualized_realtime_forecast_shock_decomposition.(['yr_' int2str(yr)]) = z(:,:,end-nfrcst:end);
|
||||||
if init>nfrcst
|
if init>nfrcst
|
||||||
|
@ -272,7 +272,7 @@ if realtime_ && isstruct(oo_) && isfield(oo_, 'realtime_shock_decomposition'),
|
||||||
oo_.annualized_realtime_conditional_shock_decomposition.(['yr_' int2str(yr-nfrcst)])(:,end,:) = ...
|
oo_.annualized_realtime_conditional_shock_decomposition.(['yr_' int2str(yr-nfrcst)])(:,end,:) = ...
|
||||||
oo_.annualized_realtime_shock_decomposition.pool(:,end,yr-nfrcst:end);
|
oo_.annualized_realtime_shock_decomposition.pool(:,end,yr-nfrcst:end);
|
||||||
if i==t1
|
if i==t1
|
||||||
for my_forecast_=(nfrcst-1):-1:1,
|
for my_forecast_=(nfrcst-1):-1:1
|
||||||
oo_.annualized_realtime_conditional_shock_decomposition.(['yr_' int2str(yr-my_forecast_)]) = ...
|
oo_.annualized_realtime_conditional_shock_decomposition.(['yr_' int2str(yr-my_forecast_)]) = ...
|
||||||
oo_.annualized_realtime_shock_decomposition.pool(:,:,yr-my_forecast_:yr) - ...
|
oo_.annualized_realtime_shock_decomposition.pool(:,:,yr-my_forecast_:yr) - ...
|
||||||
oo_.annualized_realtime_forecast_shock_decomposition.(['yr_' int2str(yr-my_forecast_)])(:,:,1:my_forecast_+1);
|
oo_.annualized_realtime_forecast_shock_decomposition.(['yr_' int2str(yr-my_forecast_)])(:,:,1:my_forecast_+1);
|
||||||
|
@ -293,24 +293,24 @@ if realtime_ && isstruct(oo_) && isfield(oo_, 'realtime_shock_decomposition'),
|
||||||
|
|
||||||
|
|
||||||
switch realtime_
|
switch realtime_
|
||||||
|
|
||||||
case 0
|
case 0
|
||||||
z = oo_.annualized_shock_decomposition;
|
z = oo_.annualized_shock_decomposition;
|
||||||
|
|
||||||
case 1 % realtime
|
case 1 % realtime
|
||||||
if vintage_
|
if vintage_
|
||||||
z = oo_.annualized_realtime_shock_decomposition.(['yr_' int2str(floor(vintage_/4))]);
|
z = oo_.annualized_realtime_shock_decomposition.(['yr_' int2str(floor(vintage_/4))]);
|
||||||
else
|
else
|
||||||
z = oo_.annualized_realtime_shock_decomposition.pool;
|
z = oo_.annualized_realtime_shock_decomposition.pool;
|
||||||
end
|
end
|
||||||
|
|
||||||
case 2 % conditional
|
case 2 % conditional
|
||||||
if vintage_
|
if vintage_
|
||||||
z = oo_.annualized_realtime_conditional_shock_decomposition.(['yr_' int2str(floor(vintage_/4))]);
|
z = oo_.annualized_realtime_conditional_shock_decomposition.(['yr_' int2str(floor(vintage_/4))]);
|
||||||
else
|
else
|
||||||
error();
|
error();
|
||||||
end
|
end
|
||||||
|
|
||||||
case 3 % forecast
|
case 3 % forecast
|
||||||
if vintage_
|
if vintage_
|
||||||
z = oo_.annualized_realtime_forecast_shock_decomposition.(['yr_' int2str(floor(vintage_/4))]);
|
z = oo_.annualized_realtime_forecast_shock_decomposition.(['yr_' int2str(floor(vintage_/4))]);
|
||||||
|
@ -320,15 +320,14 @@ if realtime_ && isstruct(oo_) && isfield(oo_, 'realtime_shock_decomposition'),
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
if q2a.plot ==0,
|
if q2a.plot ==0
|
||||||
i_var=1:2*nvar;
|
i_var=1:2*nvar;
|
||||||
steady_state = [steady_state_a;steady_state_ga];
|
steady_state = [steady_state_a;steady_state_ga];
|
||||||
else
|
else
|
||||||
i_var=1:nvar;
|
i_var=1:nvar;
|
||||||
if q2a.plot ==1,
|
if q2a.plot ==1
|
||||||
steady_state = steady_state_ga;
|
steady_state = steady_state_ga;
|
||||||
else
|
else
|
||||||
steady_state = steady_state_a;
|
steady_state = steady_state_a;
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
|
@ -2,48 +2,48 @@ function [InnovationVariance,AutoregressiveParameters] = autoregressive_process_
|
||||||
% This function computes the parameters of an AR(p) process from the variance and the autocorrelation function
|
% This function computes the parameters of an AR(p) process from the variance and the autocorrelation function
|
||||||
% (the first p terms) of this process.
|
% (the first p terms) of this process.
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% [1] Variance [double] scalar, variance of the variable.
|
% [1] Variance [double] scalar, variance of the variable.
|
||||||
% [2] Rho [double] p*1 vector, the autocorelation function: \rho(1), \rho(2), ..., \rho(p).
|
% [2] Rho [double] p*1 vector, the autocorelation function: \rho(1), \rho(2), ..., \rho(p).
|
||||||
% [3] p [double] scalar, the number of lags in the AR process.
|
% [3] p [double] scalar, the number of lags in the AR process.
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% [1] InnovationVariance [double] scalar, the variance of the innovation.
|
% [1] InnovationVariance [double] scalar, the variance of the innovation.
|
||||||
% [2] AutoregressiveParameters [double] p*1 vector of autoregressive parameters.
|
% [2] AutoregressiveParameters [double] p*1 vector of autoregressive parameters.
|
||||||
%
|
%
|
||||||
% NOTES
|
% NOTES
|
||||||
%
|
%
|
||||||
% The AR(p) model for {y_t} is:
|
% The AR(p) model for {y_t} is:
|
||||||
%
|
%
|
||||||
% y_t = \phi_1 * y_{t-1} + \phi_2 * y_{t-2} + ... + \phi_p * y_{t-p} + e_t
|
% y_t = \phi_1 * y_{t-1} + \phi_2 * y_{t-2} + ... + \phi_p * y_{t-p} + e_t
|
||||||
%
|
%
|
||||||
% Let \gamma(0) and \rho(1), ..., \rho(2) be the variance and the autocorrelation function of {y_t}. This function
|
% Let \gamma(0) and \rho(1), ..., \rho(2) be the variance and the autocorrelation function of {y_t}. This function
|
||||||
% compute the variance of {e_t} and the \phi_i (i=1,...,p) from the variance and the autocorrelation function of {y_t}.
|
% compute the variance of {e_t} and the \phi_i (i=1,...,p) from the variance and the autocorrelation function of {y_t}.
|
||||||
% We know that:
|
% We know that:
|
||||||
%
|
%
|
||||||
% \gamma(0) = \phi_1 \gamma(1) + ... + \phi_p \gamma(p) + \sigma^2
|
% \gamma(0) = \phi_1 \gamma(1) + ... + \phi_p \gamma(p) + \sigma^2
|
||||||
%
|
%
|
||||||
% where \sigma^2 is the variance of {e_t}. Equivalently we have:
|
% where \sigma^2 is the variance of {e_t}. Equivalently we have:
|
||||||
%
|
%
|
||||||
% \sigma^2 = \gamma(0) (1-\rho(1)\phi_1 - ... - \rho(p)\phi_p)
|
% \sigma^2 = \gamma(0) (1-\rho(1)\phi_1 - ... - \rho(p)\phi_p)
|
||||||
%
|
%
|
||||||
% We also have for any integer h>0:
|
% We also have for any integer h>0:
|
||||||
%
|
%
|
||||||
% \rho(h) = \phi_1 \rho(h-1) + ... + \phi_p \rho(h-p)
|
% \rho(h) = \phi_1 \rho(h-1) + ... + \phi_p \rho(h-p)
|
||||||
%
|
%
|
||||||
% We can write the equations for \rho(1), ..., \rho(p) using matrices. Let R be the p*p autocorelation
|
% We can write the equations for \rho(1), ..., \rho(p) using matrices. Let R be the p*p autocorelation
|
||||||
% matrix and v be the p*1 vector gathering the first p terms of the autocorrelation function. We have:
|
% matrix and v be the p*1 vector gathering the first p terms of the autocorrelation function. We have:
|
||||||
%
|
%
|
||||||
% v = R*PHI
|
% v = R*PHI
|
||||||
%
|
%
|
||||||
% where PHI is a p*1 vector with the autoregressive parameters of the AR(p) process. We can recover the autoregressive
|
% where PHI is a p*1 vector with the autoregressive parameters of the AR(p) process. We can recover the autoregressive
|
||||||
% parameters by inverting the autocorrelation matrix: PHI = inv(R)*v.
|
% parameters by inverting the autocorrelation matrix: PHI = inv(R)*v.
|
||||||
%
|
%
|
||||||
% This function first computes the vector PHI by inverting R and computes the variance of the innovation by evaluating
|
% This function first computes the vector PHI by inverting R and computes the variance of the innovation by evaluating
|
||||||
%
|
%
|
||||||
% \sigma^2 = \gamma(0)*(1-PHI'*v)
|
% \sigma^2 = \gamma(0)*(1-PHI'*v)
|
||||||
|
|
||||||
% Copyright (C) 2009 Dynare Team
|
% Copyright (C) 2009-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
|
|
@ -1,6 +1,6 @@
|
||||||
function [endogenousvariables, exogenousvariables] = backward_model_inversion(constraints, exogenousvariables, initialconditions, endo_names, exo_names, freeinnovations, DynareModel, DynareOptions, DynareOutput)
|
function [endogenousvariables, exogenousvariables] = backward_model_inversion(constraints, exogenousvariables, initialconditions, endo_names, exo_names, freeinnovations, DynareModel, DynareOptions, DynareOutput)
|
||||||
|
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% - constraints [dseries] with N constrained endogenous variables from t1 to t2.
|
% - constraints [dseries] with N constrained endogenous variables from t1 to t2.
|
||||||
% - exogenousvariables [dseries] with Q exogenous variables.
|
% - exogenousvariables [dseries] with Q exogenous variables.
|
||||||
% - initialconditions [dseries] with M endogenous variables starting before t1 (M initialcond must contain at least the state variables).
|
% - initialconditions [dseries] with M endogenous variables starting before t1 (M initialcond must contain at least the state variables).
|
||||||
|
@ -8,11 +8,11 @@ function [endogenousvariables, exogenousvariables] = backward_model_inversion(co
|
||||||
% - exo_names [cell] list of exogenous variable names.
|
% - exo_names [cell] list of exogenous variable names.
|
||||||
% - freeinstruments [cell] list of exogenous variable names used to control the constrained endogenous variables.
|
% - freeinstruments [cell] list of exogenous variable names used to control the constrained endogenous variables.
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% - endogenous [dseries]
|
% - endogenous [dseries]
|
||||||
% - exogenous [dseries]
|
% - exogenous [dseries]
|
||||||
%
|
%
|
||||||
% REMARKS
|
% REMARKS
|
||||||
|
|
||||||
% Copyright (C) 2017 Dynare Team
|
% Copyright (C) 2017 Dynare Team
|
||||||
%
|
%
|
||||||
|
|
|
@ -34,7 +34,7 @@ function DynareOutput = simul_backward_linear_model(initial_conditions, sample_s
|
||||||
%! @end deftypefn
|
%! @end deftypefn
|
||||||
%@eod:
|
%@eod:
|
||||||
|
|
||||||
% Copyright (C) 2012-2016 Dynare Team
|
% Copyright (C) 2012-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -84,9 +84,9 @@ A0inv = inv(jacob(:,jdx));
|
||||||
A1 = jacob(:,nonzeros(DynareModel.lead_lag_incidence(1,:)));
|
A1 = jacob(:,nonzeros(DynareModel.lead_lag_incidence(1,:)));
|
||||||
B = jacob(:,end-number_of_shocks+1:end);
|
B = jacob(:,end-number_of_shocks+1:end);
|
||||||
|
|
||||||
% Simulations
|
% Simulations
|
||||||
for it = 2:sample_size+1
|
for it = 2:sample_size+1
|
||||||
Y(:,it) = -A0inv*(cst + A1*Y(iy1,it-1) + B*DynareOutput.exo_simul(it,:)');
|
Y(:,it) = -A0inv*(cst + A1*Y(iy1,it-1) + B*DynareOutput.exo_simul(it,:)');
|
||||||
end
|
end
|
||||||
|
|
||||||
DynareOutput.endo_simul = Y;
|
DynareOutput.endo_simul = Y;
|
|
@ -34,7 +34,7 @@ function DynareOutput = simul_backward_model(initial_conditions, sample_size, Dy
|
||||||
%! @end deftypefn
|
%! @end deftypefn
|
||||||
%@eod:
|
%@eod:
|
||||||
|
|
||||||
% Copyright (C) 2012-2016 Dynare Team
|
% Copyright (C) 2012-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
|
|
@ -2,7 +2,7 @@ function DynareOutput = simul_backward_nonlinear_model(initial_conditions, sampl
|
||||||
|
|
||||||
% Simulates a stochastic non linear backward looking model with arbitrary precision (a deterministic solver is used).
|
% Simulates a stochastic non linear backward looking model with arbitrary precision (a deterministic solver is used).
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% - initial_conditions [double] n*1 vector, initial conditions for the endogenous variables.
|
% - initial_conditions [double] n*1 vector, initial conditions for the endogenous variables.
|
||||||
% - sample_size [integer] scalar, number of periods for the simulation.
|
% - sample_size [integer] scalar, number of periods for the simulation.
|
||||||
% - DynareOptions [struct] Dynare's options_ global structure.
|
% - DynareOptions [struct] Dynare's options_ global structure.
|
||||||
|
@ -10,10 +10,10 @@ function DynareOutput = simul_backward_nonlinear_model(initial_conditions, sampl
|
||||||
% - DynareOutput [struct] Dynare's oo_ global structure.
|
% - DynareOutput [struct] Dynare's oo_ global structure.
|
||||||
% - innovations [double] T*q matrix, innovations to be used for the simulation.
|
% - innovations [double] T*q matrix, innovations to be used for the simulation.
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% - DynareOutput [struct] Dynare's oo_ global structure.
|
% - DynareOutput [struct] Dynare's oo_ global structure.
|
||||||
%
|
%
|
||||||
% REMARKS
|
% REMARKS
|
||||||
% [1] The innovations used for the simulation are saved in DynareOutput.exo_simul, and the resulting paths for the endogenous
|
% [1] The innovations used for the simulation are saved in DynareOutput.exo_simul, and the resulting paths for the endogenous
|
||||||
% variables are saved in DynareOutput.endo_simul.
|
% variables are saved in DynareOutput.endo_simul.
|
||||||
% [2] The last input argument is not mandatory. If absent we use random draws and rescale them with the informations provided
|
% [2] The last input argument is not mandatory. If absent we use random draws and rescale them with the informations provided
|
||||||
|
|
|
@ -14,7 +14,7 @@ function plan = basic_plan(plan, exogenous, expectation_type, date, value)
|
||||||
% plan [structure] Returns a structure containing the updated forecast scenario.
|
% plan [structure] Returns a structure containing the updated forecast scenario.
|
||||||
%
|
%
|
||||||
%
|
%
|
||||||
% Copyright (C) 2013-2014 Dynare Team
|
% Copyright (C) 2013-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -38,7 +38,7 @@ exogenous = strtrim(exogenous);
|
||||||
ix = find(strcmp(exogenous, plan.exo_names));
|
ix = find(strcmp(exogenous, plan.exo_names));
|
||||||
if isempty(ix)
|
if isempty(ix)
|
||||||
error(['in basic_plan the second argument ' exogenous ' is not an exogenous variable']);
|
error(['in basic_plan the second argument ' exogenous ' is not an exogenous variable']);
|
||||||
end;
|
end
|
||||||
sdate = length(date);
|
sdate = length(date);
|
||||||
if sdate > 1
|
if sdate > 1
|
||||||
if date(1) < plan.date(1) || date(end) > plan.date(end)
|
if date(1) < plan.date(1) || date(end) > plan.date(end)
|
||||||
|
|
|
@ -4,14 +4,14 @@ function d = bksup0(c,ny,jcf,iyf,icf,periods)
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% ny: number of endogenous variables
|
% ny: number of endogenous variables
|
||||||
% jcf: variables index forward
|
% jcf: variables index forward
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% d: vector of backsubstitution results
|
% d: vector of backsubstitution results
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2003-2009 Dynare Team
|
% Copyright (C) 2003-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
|
|
@ -5,14 +5,14 @@ function d = bksup1(c,ny,jcf,iyf,periods)
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% ny: number of endogenous variables
|
% ny: number of endogenous variables
|
||||||
% jcf: variables index forward
|
% jcf: variables index forward
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% d: vector of backsubstitution results
|
% d: vector of backsubstitution results
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2003-2010 Dynare Team
|
% Copyright (C) 2003-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -40,4 +40,3 @@ for i = 2:periods
|
||||||
end
|
end
|
||||||
|
|
||||||
d = c(:,jcf) ;
|
d = c(:,jcf) ;
|
||||||
|
|
||||||
|
|
|
@ -15,7 +15,7 @@ function d1 = bksupk(ny,fid,jcf,icc1)
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2003-2011 Dynare Team
|
% Copyright (C) 2003-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -68,7 +68,7 @@ while i <= options_.periods
|
||||||
c = fread(fid,[jcf,ny],'float64')' ;
|
c = fread(fid,[jcf,ny],'float64')' ;
|
||||||
|
|
||||||
d1(ir) = c(:,jcf)-c(:,icf)*d1(irf) ;
|
d1(ir) = c(:,jcf)-c(:,icf)*d1(irf) ;
|
||||||
ir = ir-ny ;
|
ir = ir-ny ;
|
||||||
irf = irf-ny ;
|
irf = irf-ny ;
|
||||||
i = i+1;
|
i = i+1;
|
||||||
end
|
end
|
||||||
|
|
|
@ -1,6 +1,6 @@
|
||||||
function x = bseastr(s1,s2)
|
function x = bseastr(s1,s2)
|
||||||
|
|
||||||
% Copyright (C) 2001-2009 Dynare Team
|
% Copyright (C) 2001-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -33,16 +33,15 @@ for im = 1:m
|
||||||
for i = 1:min(length(key),length(temp))
|
for i = 1:min(length(key),length(temp))
|
||||||
if temp(i) > key(i)
|
if temp(i) > key(i)
|
||||||
h = mid - 1 ;
|
h = mid - 1 ;
|
||||||
break
|
break
|
||||||
else
|
else
|
||||||
l = mid + 1 ;
|
l = mid + 1 ;
|
||||||
break
|
break
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
else
|
else
|
||||||
x(im) = mid ;
|
x(im) = mid ;
|
||||||
break
|
break
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
|
@ -12,7 +12,7 @@ function bvar_density(maxnlags)
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2003-2007 Christopher Sims
|
% Copyright (C) 2003-2007 Christopher Sims
|
||||||
% Copyright (C) 2007-2016 Dynare Team
|
% Copyright (C) 2007-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -37,16 +37,16 @@ for nlags = 1:maxnlags
|
||||||
[ny, nx, posterior, prior] = bvar_toolbox(nlags);
|
[ny, nx, posterior, prior] = bvar_toolbox(nlags);
|
||||||
oo_.bvar.posterior{nlags}=posterior;
|
oo_.bvar.posterior{nlags}=posterior;
|
||||||
oo_.bvar.prior{nlags}=prior;
|
oo_.bvar.prior{nlags}=prior;
|
||||||
|
|
||||||
posterior_int = matrictint(posterior.S, posterior.df, posterior.XXi);
|
posterior_int = matrictint(posterior.S, posterior.df, posterior.XXi);
|
||||||
prior_int = matrictint(prior.S, prior.df, prior.XXi);
|
prior_int = matrictint(prior.S, prior.df, prior.XXi);
|
||||||
|
|
||||||
lik_nobs = posterior.df - prior.df;
|
lik_nobs = posterior.df - prior.df;
|
||||||
|
|
||||||
log_dnsty = posterior_int - prior_int - 0.5*ny*lik_nobs*log(2*pi);
|
log_dnsty = posterior_int - prior_int - 0.5*ny*lik_nobs*log(2*pi);
|
||||||
|
|
||||||
oo_.bvar.log_marginal_data_density(nlags)=log_dnsty;
|
oo_.bvar.log_marginal_data_density(nlags)=log_dnsty;
|
||||||
|
|
||||||
skipline()
|
skipline()
|
||||||
fprintf('The marginal log density of the BVAR(%g) model is equal to %10.4f\n', ...
|
fprintf('The marginal log density of the BVAR(%g) model is equal to %10.4f\n', ...
|
||||||
nlags, log_dnsty);
|
nlags, log_dnsty);
|
||||||
|
@ -61,7 +61,7 @@ function w = matrictint(S, df, XXi)
|
||||||
% S: parameter of inverse-Wishart distribution
|
% S: parameter of inverse-Wishart distribution
|
||||||
% df: number of degrees of freedom of inverse-Wishart distribution
|
% df: number of degrees of freedom of inverse-Wishart distribution
|
||||||
% XXi: first component of VCV matrix of matrix-normal distribution
|
% XXi: first component of VCV matrix of matrix-normal distribution
|
||||||
%
|
%
|
||||||
% Computes the integral over (Phi, Sigma) of:
|
% Computes the integral over (Phi, Sigma) of:
|
||||||
%
|
%
|
||||||
% det(Sigma)^(-k/2)*exp(-0.5*Tr((Phi-PhiHat)'*(XXi)^(-1)*(Phi-PhiHat)*Sigma^(-1)))*
|
% det(Sigma)^(-k/2)*exp(-0.5*Tr((Phi-PhiHat)'*(XXi)^(-1)*(Phi-PhiHat)*Sigma^(-1)))*
|
||||||
|
|
|
@ -54,7 +54,7 @@ p = 0;
|
||||||
% Loop counter initialization
|
% Loop counter initialization
|
||||||
d = 0;
|
d = 0;
|
||||||
while d <= options_.bvar_replic
|
while d <= options_.bvar_replic
|
||||||
|
|
||||||
Sigma = rand_inverse_wishart(ny, posterior.df, S_inv_upper_chol);
|
Sigma = rand_inverse_wishart(ny, posterior.df, S_inv_upper_chol);
|
||||||
|
|
||||||
% Option 'lower' of chol() not available in old versions of
|
% Option 'lower' of chol() not available in old versions of
|
||||||
|
@ -62,15 +62,15 @@ while d <= options_.bvar_replic
|
||||||
Sigma_lower_chol = chol(Sigma)';
|
Sigma_lower_chol = chol(Sigma)';
|
||||||
|
|
||||||
Phi = rand_matrix_normal(k, ny, posterior.PhiHat, Sigma_lower_chol, XXi_lower_chol);
|
Phi = rand_matrix_normal(k, ny, posterior.PhiHat, Sigma_lower_chol, XXi_lower_chol);
|
||||||
|
|
||||||
% All the eigenvalues of the companion matrix have to be on or inside the unit circle
|
% All the eigenvalues of the companion matrix have to be on or inside the unit circle
|
||||||
Companion_matrix(1:ny,:) = Phi(1:ny*nlags,:)';
|
Companion_matrix(1:ny,:) = Phi(1:ny*nlags,:)';
|
||||||
test = (abs(eig(Companion_matrix)));
|
test = (abs(eig(Companion_matrix)));
|
||||||
if any(test>1.0000000000001)
|
if any(test>1.0000000000001)
|
||||||
p = p+1;
|
p = p+1;
|
||||||
end
|
end
|
||||||
d = d+1;
|
d = d+1;
|
||||||
|
|
||||||
% Without shocks
|
% Without shocks
|
||||||
lags_data = forecast_data.initval;
|
lags_data = forecast_data.initval;
|
||||||
for t = 1:options_.forecast
|
for t = 1:options_.forecast
|
||||||
|
@ -80,7 +80,7 @@ while d <= options_.bvar_replic
|
||||||
lags_data(end,:) = y;
|
lags_data(end,:) = y;
|
||||||
sims_no_shock(t, :, d) = y;
|
sims_no_shock(t, :, d) = y;
|
||||||
end
|
end
|
||||||
|
|
||||||
% With shocks
|
% With shocks
|
||||||
lags_data = forecast_data.initval;
|
lags_data = forecast_data.initval;
|
||||||
for t = 1:options_.forecast
|
for t = 1:options_.forecast
|
||||||
|
@ -130,9 +130,9 @@ dyn_saveas(dyn_graph.fh,[OutputDirectoryName '/' M_.fname '_BVAR_forecast_',num2
|
||||||
% Compute RMSE
|
% Compute RMSE
|
||||||
|
|
||||||
if ~isempty(forecast_data.realized_val)
|
if ~isempty(forecast_data.realized_val)
|
||||||
|
|
||||||
sq_err_cumul = zeros(1, ny);
|
sq_err_cumul = zeros(1, ny);
|
||||||
|
|
||||||
lags_data = forecast_data.initval;
|
lags_data = forecast_data.initval;
|
||||||
for t = 1:size(forecast_data.realized_val, 1)
|
for t = 1:size(forecast_data.realized_val, 1)
|
||||||
X = [ reshape(flipdim(lags_data, 1)', 1, ny*nlags) forecast_data.realized_xdata(t, :) ];
|
X = [ reshape(flipdim(lags_data, 1)', 1, ny*nlags) forecast_data.realized_xdata(t, :) ];
|
||||||
|
@ -141,14 +141,14 @@ if ~isempty(forecast_data.realized_val)
|
||||||
lags_data(end,:) = y;
|
lags_data(end,:) = y;
|
||||||
sq_err_cumul = sq_err_cumul + (y - forecast_data.realized_val(t, :)) .^ 2;
|
sq_err_cumul = sq_err_cumul + (y - forecast_data.realized_val(t, :)) .^ 2;
|
||||||
end
|
end
|
||||||
|
|
||||||
rmse = sqrt(sq_err_cumul / size(forecast_data.realized_val, 1));
|
rmse = sqrt(sq_err_cumul / size(forecast_data.realized_val, 1));
|
||||||
|
|
||||||
fprintf('RMSE of BVAR(%d):\n', nlags);
|
fprintf('RMSE of BVAR(%d):\n', nlags);
|
||||||
|
|
||||||
for i = 1:length(options_.varobs)
|
for i = 1:length(options_.varobs)
|
||||||
fprintf('%s: %10.4f\n', options_.varobs{i}, rmse(i));
|
fprintf('%s: %10.4f\n', options_.varobs{i}, rmse(i));
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
% Store results
|
% Store results
|
||||||
|
|
|
@ -11,7 +11,7 @@ function bvar_irf(nlags,identification)
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2007-2012 Dynare Team
|
% Copyright (C) 2007-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -54,7 +54,7 @@ p = 0;
|
||||||
sampled_irfs = NaN(ny, ny, options_.irf, options_.bvar_replic);
|
sampled_irfs = NaN(ny, ny, options_.irf, options_.bvar_replic);
|
||||||
|
|
||||||
for draw=1:options_.bvar_replic
|
for draw=1:options_.bvar_replic
|
||||||
|
|
||||||
% Get a covariance matrix from an inverted Wishart distribution.
|
% Get a covariance matrix from an inverted Wishart distribution.
|
||||||
Sigma = rand_inverse_wishart(ny, posterior.df, S_inv_upper_chol);
|
Sigma = rand_inverse_wishart(ny, posterior.df, S_inv_upper_chol);
|
||||||
Sigma_upper_chol = chol(Sigma);
|
Sigma_upper_chol = chol(Sigma);
|
||||||
|
@ -62,10 +62,10 @@ for draw=1:options_.bvar_replic
|
||||||
|
|
||||||
% Get the Autoregressive matrices from a matrix variate distribution.
|
% Get the Autoregressive matrices from a matrix variate distribution.
|
||||||
Phi = rand_matrix_normal(k, ny, posterior.PhiHat, Sigma_lower_chol, XXi_lower_chol);
|
Phi = rand_matrix_normal(k, ny, posterior.PhiHat, Sigma_lower_chol, XXi_lower_chol);
|
||||||
|
|
||||||
% Form the companion matrix.
|
% Form the companion matrix.
|
||||||
Companion_matrix(1:ny,:) = transpose(Phi(1:ny*nlags,:));
|
Companion_matrix(1:ny,:) = transpose(Phi(1:ny*nlags,:));
|
||||||
|
|
||||||
% All the eigenvalues of the companion matrix have to be on or
|
% All the eigenvalues of the companion matrix have to be on or
|
||||||
% inside the unit circle to rule out explosive time series.
|
% inside the unit circle to rule out explosive time series.
|
||||||
test = (abs(eig(Companion_matrix)));
|
test = (abs(eig(Companion_matrix)));
|
||||||
|
@ -78,7 +78,7 @@ for draw=1:options_.bvar_replic
|
||||||
elseif strcmpi(identification,'SquareRoot')
|
elseif strcmpi(identification,'SquareRoot')
|
||||||
StructuralMat = sqrtm(Sigma);
|
StructuralMat = sqrtm(Sigma);
|
||||||
end
|
end
|
||||||
|
|
||||||
% Build the IRFs...
|
% Build the IRFs...
|
||||||
lags_data = zeros(ny,ny*nlags) ;
|
lags_data = zeros(ny,ny*nlags) ;
|
||||||
sampled_irfs(:,:,1,draw) = Sigma_lower_chol ;
|
sampled_irfs(:,:,1,draw) = Sigma_lower_chol ;
|
||||||
|
@ -88,7 +88,7 @@ for draw=1:options_.bvar_replic
|
||||||
lags_data(:,ny+1:end) = lags_data(:,1:end-ny) ;
|
lags_data(:,ny+1:end) = lags_data(:,1:end-ny) ;
|
||||||
lags_data(:,1:ny) = sampled_irfs(:,:,t,draw) ;
|
lags_data(:,1:ny) = sampled_irfs(:,:,t,draw) ;
|
||||||
end
|
end
|
||||||
|
|
||||||
end
|
end
|
||||||
|
|
||||||
if p > 0
|
if p > 0
|
||||||
|
@ -106,7 +106,7 @@ sort_idx = round((0.5 + [-options_.bvar.conf_sig, options_.bvar.conf_sig, .0]/2)
|
||||||
|
|
||||||
posterior_down_conf_irfs = sorted_irfs(:,:,:,sort_idx(1));
|
posterior_down_conf_irfs = sorted_irfs(:,:,:,sort_idx(1));
|
||||||
posterior_up_conf_irfs = sorted_irfs(:,:,:,sort_idx(2));
|
posterior_up_conf_irfs = sorted_irfs(:,:,:,sort_idx(2));
|
||||||
posterior_median_irfs = sorted_irfs(:,:,:,sort_idx(3));
|
posterior_median_irfs = sorted_irfs(:,:,:,sort_idx(3));
|
||||||
|
|
||||||
number_of_columns = fix(sqrt(ny));
|
number_of_columns = fix(sqrt(ny));
|
||||||
number_of_rows = ceil(ny / number_of_columns) ;
|
number_of_rows = ceil(ny / number_of_columns) ;
|
||||||
|
|
|
@ -29,7 +29,7 @@ function [ny, nx, posterior, prior, forecast_data] = bvar_toolbox(nlags)
|
||||||
% forecasting, of size options_.forecast*nx (actually only
|
% forecasting, of size options_.forecast*nx (actually only
|
||||||
% contains "1" values for the constant term if nx ~= 0)
|
% contains "1" values for the constant term if nx ~= 0)
|
||||||
% - realized_val: only non-empty if options_.nobs doesn't point
|
% - realized_val: only non-empty if options_.nobs doesn't point
|
||||||
% to the end of sample
|
% to the end of sample
|
||||||
% In that case, contains values of endogenous variables after
|
% In that case, contains values of endogenous variables after
|
||||||
% options_.nobs and up to the end of the sample
|
% options_.nobs and up to the end of the sample
|
||||||
% - realized_xdata: contains values of exogenous variables after
|
% - realized_xdata: contains values of exogenous variables after
|
||||||
|
@ -42,7 +42,7 @@ function [ny, nx, posterior, prior, forecast_data] = bvar_toolbox(nlags)
|
||||||
% - bvar_prior_{tau,decay,lambda,mu,omega,flat,train}
|
% - bvar_prior_{tau,decay,lambda,mu,omega,flat,train}
|
||||||
|
|
||||||
% Copyright (C) 2003-2007 Christopher Sims
|
% Copyright (C) 2003-2007 Christopher Sims
|
||||||
% Copyright (C) 2007-2012 Dynare Team
|
% Copyright (C) 2007-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -67,7 +67,7 @@ options_ = set_default_option(options_, 'nobs', size(dataset,1)-options_.first_o
|
||||||
|
|
||||||
if (options_.first_obs+options_.nobs-1)> size(dataset,1)
|
if (options_.first_obs+options_.nobs-1)> size(dataset,1)
|
||||||
fprintf('Incorrect or missing specification of the number of observations. nobs can be at most %4u\n',size(dataset,1)-options_.first_obs+1);
|
fprintf('Incorrect or missing specification of the number of observations. nobs can be at most %4u\n',size(dataset,1)-options_.first_obs+1);
|
||||||
error('Inconsistent number of observations.')
|
error('Inconsistent number of observations.')
|
||||||
end
|
end
|
||||||
|
|
||||||
% Parameters for prior
|
% Parameters for prior
|
||||||
|
@ -160,7 +160,7 @@ function [ydum,xdum,breaks]=varprior(nv,nx,lags,mnprior,vprior)
|
||||||
%function [ydum,xdum,breaks]=varprior(nv,nx,lags,mnprior,vprior)
|
%function [ydum,xdum,breaks]=varprior(nv,nx,lags,mnprior,vprior)
|
||||||
% ydum, xdum: dummy observation data that implement the prior
|
% ydum, xdum: dummy observation data that implement the prior
|
||||||
% breaks: vector of points in the dummy data after which new dummy obs's start
|
% breaks: vector of points in the dummy data after which new dummy obs's start
|
||||||
% Set breaks=T+[0;breaks], ydata=[ydata;ydum], xdum=[xdata;xdum], where
|
% Set breaks=T+[0;breaks], ydata=[ydata;ydum], xdum=[xdata;xdum], where
|
||||||
% actual data matrix has T rows, in preparing input for rfvar3
|
% actual data matrix has T rows, in preparing input for rfvar3
|
||||||
% nv,nx,lags: VAR dimensions
|
% nv,nx,lags: VAR dimensions
|
||||||
% mnprior.tight:Overall tightness of Minnesota prior
|
% mnprior.tight:Overall tightness of Minnesota prior
|
||||||
|
@ -175,8 +175,8 @@ function [ydum,xdum,breaks]=varprior(nv,nx,lags,mnprior,vprior)
|
||||||
% taken to include the sum-of-coefficients and co-persistence components
|
% taken to include the sum-of-coefficients and co-persistence components
|
||||||
% that are implemented directly in rfvar3.m. The diagonal prior on v, combined
|
% that are implemented directly in rfvar3.m. The diagonal prior on v, combined
|
||||||
% with sum-of-coefficients and co-persistence components and with the unit own-first-lag
|
% with sum-of-coefficients and co-persistence components and with the unit own-first-lag
|
||||||
% prior mean generates larger prior variances for own than for cross-effects even in
|
% prior mean generates larger prior variances for own than for cross-effects even in
|
||||||
% this formulation, but here there is no way to shrink toward a set of unconstrained
|
% this formulation, but here there is no way to shrink toward a set of unconstrained
|
||||||
% univariate AR's.
|
% univariate AR's.
|
||||||
|
|
||||||
% Original file downloaded from:
|
% Original file downloaded from:
|
||||||
|
@ -216,3 +216,108 @@ dimy = size(ydum);
|
||||||
ydum = reshape(permute(ydum,[1 3 2]),dimy(1)*dimy(3),nv);
|
ydum = reshape(permute(ydum,[1 3 2]),dimy(1)*dimy(3),nv);
|
||||||
xdum = reshape(permute(xdum,[1 3 2]),dimy(1)*dimy(3),nx);
|
xdum = reshape(permute(xdum,[1 3 2]),dimy(1)*dimy(3),nx);
|
||||||
breaks = breaks(1:(end-1));
|
breaks = breaks(1:(end-1));
|
||||||
|
|
||||||
|
|
||||||
|
function var=rfvar3(ydata,lags,xdata,breaks,lambda,mu)
|
||||||
|
%function var=rfvar3(ydata,lags,xdata,breaks,lambda,mu)
|
||||||
|
% This algorithm goes for accuracy without worrying about memory requirements.
|
||||||
|
% ydata: dependent variable data matrix
|
||||||
|
% xdata: exogenous variable data matrix
|
||||||
|
% lags: number of lags
|
||||||
|
% breaks: rows in ydata and xdata after which there is a break. This allows for
|
||||||
|
% discontinuities in the data (e.g. war years) and for the possibility of
|
||||||
|
% adding dummy observations to implement a prior. This must be a column vector.
|
||||||
|
% Note that a single dummy observation becomes lags+1 rows of the data matrix,
|
||||||
|
% with a break separating it from the rest of the data. The function treats the
|
||||||
|
% first lags observations at the top and after each "break" in ydata and xdata as
|
||||||
|
% initial conditions.
|
||||||
|
% lambda: weight on "co-persistence" prior dummy observations. This expresses
|
||||||
|
% belief that when data on *all* y's are stable at their initial levels, they will
|
||||||
|
% tend to persist at that level. lambda=5 is a reasonable first try. With lambda<0,
|
||||||
|
% constant term is not included in the dummy observation, so that stationary models
|
||||||
|
% with means equal to initial ybar do not fit the prior mean. With lambda>0, the prior
|
||||||
|
% implies that large constants are unlikely if unit roots are present.
|
||||||
|
% mu: weight on "own persistence" prior dummy observation. Expresses belief
|
||||||
|
% that when y_i has been stable at its initial level, it will tend to persist
|
||||||
|
% at that level, regardless of the values of other variables. There is
|
||||||
|
% one of these for each variable. A reasonable first guess is mu=2.
|
||||||
|
% The program assumes that the first lags rows of ydata and xdata are real data, not dummies.
|
||||||
|
% Dummy observations should go at the end, if any. If pre-sample x's are not available,
|
||||||
|
% repeating the initial xdata(lags+1,:) row or copying xdata(lags+1:2*lags,:) into
|
||||||
|
% xdata(1:lags,:) are reasonable subsititutes. These values are used in forming the
|
||||||
|
% persistence priors.
|
||||||
|
|
||||||
|
% Original file downloaded from:
|
||||||
|
% http://sims.princeton.edu/yftp/VARtools/matlab/rfvar3.m
|
||||||
|
|
||||||
|
[T,nvar] = size(ydata);
|
||||||
|
nox = isempty(xdata);
|
||||||
|
if ~nox
|
||||||
|
[T2,nx] = size(xdata);
|
||||||
|
else
|
||||||
|
T2 = T;
|
||||||
|
nx = 0;
|
||||||
|
xdata = zeros(T2,0);
|
||||||
|
end
|
||||||
|
% note that x must be same length as y, even though first part of x will not be used.
|
||||||
|
% This is so that the lags parameter can be changed without reshaping the xdata matrix.
|
||||||
|
if T2 ~= T, error('Mismatch of x and y data lengths'),end
|
||||||
|
if nargin < 4
|
||||||
|
nbreaks = 0;
|
||||||
|
breaks = [];
|
||||||
|
else
|
||||||
|
nbreaks = length(breaks);
|
||||||
|
end
|
||||||
|
breaks = [0;breaks;T];
|
||||||
|
smpl = [];
|
||||||
|
for nb = 1:nbreaks+1
|
||||||
|
smpl = [smpl;[breaks(nb)+lags+1:breaks(nb+1)]'];
|
||||||
|
end
|
||||||
|
Tsmpl = size(smpl,1);
|
||||||
|
X = zeros(Tsmpl,nvar,lags);
|
||||||
|
for is = 1:length(smpl)
|
||||||
|
X(is,:,:) = ydata(smpl(is)-(1:lags),:)';
|
||||||
|
end
|
||||||
|
X = [X(:,:) xdata(smpl,:)];
|
||||||
|
y = ydata(smpl,:);
|
||||||
|
% Everything now set up with input data for y=Xb+e
|
||||||
|
|
||||||
|
% Add persistence dummies
|
||||||
|
if lambda ~= 0 || mu > 0
|
||||||
|
ybar = mean(ydata(1:lags,:),1);
|
||||||
|
if ~nox
|
||||||
|
xbar = mean(xdata(1:lags,:),1);
|
||||||
|
else
|
||||||
|
xbar = [];
|
||||||
|
end
|
||||||
|
if lambda ~= 0
|
||||||
|
if lambda>0
|
||||||
|
xdum = lambda*[repmat(ybar,1,lags) xbar];
|
||||||
|
else
|
||||||
|
lambda = -lambda;
|
||||||
|
xdum = lambda*[repmat(ybar,1,lags) zeros(size(xbar))];
|
||||||
|
end
|
||||||
|
ydum = zeros(1,nvar);
|
||||||
|
ydum(1,:) = lambda*ybar;
|
||||||
|
y = [y;ydum];
|
||||||
|
X = [X;xdum];
|
||||||
|
end
|
||||||
|
if mu>0
|
||||||
|
xdum = [repmat(diag(ybar),1,lags) zeros(nvar,nx)]*mu;
|
||||||
|
ydum = mu*diag(ybar);
|
||||||
|
X = [X;xdum];
|
||||||
|
y = [y;ydum];
|
||||||
|
end
|
||||||
|
end
|
||||||
|
|
||||||
|
% Compute OLS regression and residuals
|
||||||
|
[vl,d,vr] = svd(X,0);
|
||||||
|
di = 1./diag(d);
|
||||||
|
B = (vr.*repmat(di',nvar*lags+nx,1))*vl'*y;
|
||||||
|
u = y-X*B;
|
||||||
|
xxi = vr.*repmat(di',nvar*lags+nx,1);
|
||||||
|
xxi = xxi*xxi';
|
||||||
|
|
||||||
|
var.B = B;
|
||||||
|
var.u = u;
|
||||||
|
var.xxi = xxi;
|
||||||
|
|
|
@ -5,7 +5,7 @@ function cprod = cartesian_product_of_sets(varargin)
|
||||||
%! @deftypefn {Function File} {@var{cprod} =} cartesian_product_of_sets (@var{a},@var{b}, ...)
|
%! @deftypefn {Function File} {@var{cprod} =} cartesian_product_of_sets (@var{a},@var{b}, ...)
|
||||||
%! @anchor{cartesian_product_of_sets}
|
%! @anchor{cartesian_product_of_sets}
|
||||||
%! @sp 1
|
%! @sp 1
|
||||||
%! Computes A_1 * A_2 * .... * A_n with a generic set A_i = {e_1,e_2,e_3,...} where e_i is a string
|
%! Computes A_1 * A_2 * .... * A_n with a generic set A_i = {e_1,e_2,e_3,...} where e_i is a string
|
||||||
%! or a number. It is assumed that each element e_i is unique in set A_i.
|
%! or a number. It is assumed that each element e_i is unique in set A_i.
|
||||||
%! @sp 2
|
%! @sp 2
|
||||||
%! @strong{Inputs}
|
%! @strong{Inputs}
|
||||||
|
@ -31,7 +31,7 @@ function cprod = cartesian_product_of_sets(varargin)
|
||||||
%! @end deftypefn
|
%! @end deftypefn
|
||||||
%@eod:
|
%@eod:
|
||||||
|
|
||||||
% Copyright (C) 2011-2012 Dynare Team
|
% Copyright (C) 2011-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
|
|
@ -2,14 +2,14 @@ function cellofchar2mfile(fname, c, cname)
|
||||||
|
|
||||||
% Write a cell of char in a matlab script.
|
% Write a cell of char in a matlab script.
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% - fname [string] name of the file where c is to be saved.
|
% - fname [string] name of the file where c is to be saved.
|
||||||
% - c [cell] a two dimensional cell of char.
|
% - c [cell] a two dimensional cell of char.
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% None.
|
% None.
|
||||||
|
|
||||||
% Copyright (C) 2015 Dynare Team
|
% Copyright (C) 2015-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
|
|
@ -40,7 +40,7 @@ function [eigenvalues_,result,info] = check(M, options, oo)
|
||||||
%! @end deftypefn
|
%! @end deftypefn
|
||||||
%@eod:
|
%@eod:
|
||||||
|
|
||||||
% Copyright (C) 2001-2013 Dynare Team
|
% Copyright (C) 2001-2014 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
|
|
@ -2,7 +2,7 @@ function check_dsge_var_model(Model, EstimatedParameters, BayesInfo)
|
||||||
|
|
||||||
% Check if the dsge model can be estimated with the DSGE-VAR approach.
|
% Check if the dsge model can be estimated with the DSGE-VAR approach.
|
||||||
|
|
||||||
% Copyright (C) 2013 Dynare Team
|
% Copyright (C) 2013-2014 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
|
|
@ -12,7 +12,7 @@ function estim_params=check_for_calibrated_covariances(xparam1,estim_params,M)
|
||||||
% Notes: M is local to this function and not updated when calling
|
% Notes: M is local to this function and not updated when calling
|
||||||
% set_all_parameters
|
% set_all_parameters
|
||||||
|
|
||||||
% Copyright (C) 2013 Dynare Team
|
% Copyright (C) 2013-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -42,7 +42,7 @@ calibrated_covariance_pos=covariance_pos(~ismember(covariance_pos,correlation_po
|
||||||
if any(calibrated_covariance_pos)
|
if any(calibrated_covariance_pos)
|
||||||
[rows, columns]=ind2sub(size(M.Sigma_e),calibrated_covariance_pos); %find linear indices of lower triangular covariance entries
|
[rows, columns]=ind2sub(size(M.Sigma_e),calibrated_covariance_pos); %find linear indices of lower triangular covariance entries
|
||||||
estim_params.calibrated_covariances.position=[calibrated_covariance_pos;sub2ind(size(M.Sigma_e),columns,rows)]; %get linear entries of upper triangular parts
|
estim_params.calibrated_covariances.position=[calibrated_covariance_pos;sub2ind(size(M.Sigma_e),columns,rows)]; %get linear entries of upper triangular parts
|
||||||
estim_params.calibrated_covariances.cov_value=Sigma_e_calibrated(estim_params.calibrated_covariances.position);
|
estim_params.calibrated_covariances.cov_value=Sigma_e_calibrated(estim_params.calibrated_covariances.position);
|
||||||
end
|
end
|
||||||
|
|
||||||
correlation_pos_ME=find(tril(M.Correlation_matrix_ME,-1)); %off-diagonal elements set by correlations after accounting for estimation
|
correlation_pos_ME=find(tril(M.Correlation_matrix_ME,-1)); %off-diagonal elements set by correlations after accounting for estimation
|
||||||
|
@ -50,6 +50,5 @@ calibrated_covariance_pos_ME=covariance_pos_ME(~ismember(covariance_pos_ME,corre
|
||||||
if any(calibrated_covariance_pos_ME)
|
if any(calibrated_covariance_pos_ME)
|
||||||
[rows, columns]=ind2sub(size(M.H),calibrated_covariance_pos_ME); %find linear indices of lower triangular covariance entries
|
[rows, columns]=ind2sub(size(M.H),calibrated_covariance_pos_ME); %find linear indices of lower triangular covariance entries
|
||||||
estim_params.calibrated_covariances_ME.position=[calibrated_covariance_pos_ME;sub2ind(size(M.H),columns,rows)]; %get linear entries of upper triangular parts
|
estim_params.calibrated_covariances_ME.position=[calibrated_covariance_pos_ME;sub2ind(size(M.H),columns,rows)]; %get linear entries of upper triangular parts
|
||||||
estim_params.calibrated_covariances_ME.cov_value=H_calibrated(estim_params.calibrated_covariances_ME.position);
|
estim_params.calibrated_covariances_ME.cov_value=H_calibrated(estim_params.calibrated_covariances_ME.position);
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
|
@ -1,20 +1,20 @@
|
||||||
function varlist = check_list_of_variables(options_, M_, varlist)
|
function varlist = check_list_of_variables(options_, M_, varlist)
|
||||||
% This function defines, if necessary, the list of endogenous variables
|
% This function defines, if necessary, the list of endogenous variables
|
||||||
% for which the posterior statistics have to be computed.
|
% for which the posterior statistics have to be computed.
|
||||||
%
|
|
||||||
%
|
%
|
||||||
% INPUTS
|
%
|
||||||
|
% INPUTS
|
||||||
%
|
%
|
||||||
% options_ [structure] Dynare structure.
|
% options_ [structure] Dynare structure.
|
||||||
% M_ [structure] Dynare structure (related to model definition).
|
% M_ [structure] Dynare structure (related to model definition).
|
||||||
% varlist [string] Array of strings with name of the endogenous variables.
|
% varlist [string] Array of strings with name of the endogenous variables.
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% varlist [string]
|
% varlist [string]
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
|
|
||||||
% Copyright (C) 2003-2014 Dynare Team
|
% Copyright (C) 2003-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -33,8 +33,8 @@ function varlist = check_list_of_variables(options_, M_, varlist)
|
||||||
|
|
||||||
%get uniques
|
%get uniques
|
||||||
|
|
||||||
[junk1,junk2,index_uniqes] = varlist_indices(varlist,M_.endo_names);
|
[junk1,junk2,index_uniques] = varlist_indices(varlist,M_.endo_names);
|
||||||
varlist=varlist(index_uniqes,:);
|
varlist=varlist(index_uniques,:);
|
||||||
|
|
||||||
msg = 0;
|
msg = 0;
|
||||||
if options_.dsge_var && options_.bayesian_irf
|
if options_.dsge_var && options_.bayesian_irf
|
||||||
|
@ -63,7 +63,7 @@ if ~isempty(varlist) && ~isempty(options_.endo_vars_for_moment_computations_in_e
|
||||||
error('You cannot use the consider_all_endogenous or consider_all_observed options when listing variables after the estimation command')
|
error('You cannot use the consider_all_endogenous or consider_all_observed options when listing variables after the estimation command')
|
||||||
elseif isempty(varlist) && ~isempty(options_.endo_vars_for_moment_computations_in_estimation)
|
elseif isempty(varlist) && ~isempty(options_.endo_vars_for_moment_computations_in_estimation)
|
||||||
if strcmp(options_.endo_vars_for_moment_computations_in_estimation,'all_endogenous_variables')
|
if strcmp(options_.endo_vars_for_moment_computations_in_estimation,'all_endogenous_variables')
|
||||||
varlist = M_.endo_names(1:M_.orig_endo_nbr, :);
|
varlist = M_.endo_names(1:M_.orig_endo_nbr, :);
|
||||||
elseif strcmp(options_.endo_vars_for_moment_computations_in_estimation,'only_observed_variables')
|
elseif strcmp(options_.endo_vars_for_moment_computations_in_estimation,'only_observed_variables')
|
||||||
varlist = char(options_.varobs');
|
varlist = char(options_.varobs');
|
||||||
else
|
else
|
||||||
|
@ -106,7 +106,7 @@ elseif isempty(varlist) && isempty(options_.endo_vars_for_moment_computations_in
|
||||||
end
|
end
|
||||||
if ~isempty(cas)
|
if ~isempty(cas)
|
||||||
string = [ cas , ' will be computed for the ' num2str(M_.endo_nbr) ' endogenous variables'];
|
string = [ cas , ' will be computed for the ' num2str(M_.endo_nbr) ' endogenous variables'];
|
||||||
string = [ string ' of your model, this can be very long....'];
|
string = [ string ' of your model, this can be very long....'];
|
||||||
format_text(string, 10)
|
format_text(string, 10)
|
||||||
if options_.nointeractive
|
if options_.nointeractive
|
||||||
% Default behaviour is to consider all the endogenous variables.
|
% Default behaviour is to consider all the endogenous variables.
|
||||||
|
|
|
@ -1,6 +1,6 @@
|
||||||
function check_matlab_path(change_path_flag)
|
function check_matlab_path(change_path_flag)
|
||||||
|
|
||||||
% Copyright (C) 2015-2016 Dynare Team
|
% Copyright (C) 2015-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -82,7 +82,7 @@ else
|
||||||
warning(msg);
|
warning(msg);
|
||||||
skipline()
|
skipline()
|
||||||
rmpath(DYNARE_PATH)
|
rmpath(DYNARE_PATH)
|
||||||
addpath(DYNARE_PATH)
|
addpath(DYNARE_PATH)
|
||||||
end
|
end
|
||||||
warning on backtrace
|
warning on backtrace
|
||||||
end
|
end
|
||||||
|
@ -90,7 +90,7 @@ else
|
||||||
else
|
else
|
||||||
% Check that the user did not put all the subfolders in the path.
|
% Check that the user did not put all the subfolders in the path.
|
||||||
% => If DYNARE_PATH/qz is in the path while mjdgges dll is available
|
% => If DYNARE_PATH/qz is in the path while mjdgges dll is available
|
||||||
% it most likely means that user wrongly put all subfolders in the
|
% it most likely means that user wrongly put all subfolders in the
|
||||||
% matlab's path!
|
% matlab's path!
|
||||||
mexpath = add_path_to_mex_files([DYNARE_PATH filesep], false);
|
mexpath = add_path_to_mex_files([DYNARE_PATH filesep], false);
|
||||||
MATLAB_PATH = path2cell(MATLAB_PATH);
|
MATLAB_PATH = path2cell(MATLAB_PATH);
|
||||||
|
@ -107,14 +107,14 @@ else
|
||||||
end
|
end
|
||||||
|
|
||||||
function q = path2cell(p)
|
function q = path2cell(p)
|
||||||
% Converts the output of path() to a cell
|
% Converts the output of path() to a cell
|
||||||
s = strfind(p,pathsep);
|
s = strfind(p,pathsep);
|
||||||
n = length(s)+1;
|
n = length(s)+1;
|
||||||
q = cell(n,1);
|
q = cell(n,1);
|
||||||
q(1) = {p(1:s(1)-1)};
|
q(1) = {p(1:s(1)-1)};
|
||||||
q(n) = {p(s(end)+1:end)};
|
q(n) = {p(s(end)+1:end)};
|
||||||
for i=2:n-1
|
for i=2:n-1
|
||||||
q(i) = {p(s(i-1)+1:s(i)-1)};
|
q(i) = {p(s(i-1)+1:s(i)-1)};
|
||||||
end
|
end
|
||||||
|
|
||||||
function flist = getallroutinenames(p, excludedsubfolders)
|
function flist = getallroutinenames(p, excludedsubfolders)
|
||||||
|
|
|
@ -1,6 +1,6 @@
|
||||||
function check_model(DynareModel)
|
function check_model(DynareModel)
|
||||||
|
|
||||||
% Copyright (C) 2005-2012 Dynare Team
|
% Copyright (C) 2005-2013 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
|
|
@ -1,13 +1,13 @@
|
||||||
function [info,description] = check_posterior_analysis_data(type,M_)
|
function [info,description] = check_posterior_analysis_data(type,M_)
|
||||||
% function [info,description] = check_posterior_analysis_data(type,M_)
|
% function [info,description] = check_posterior_analysis_data(type,M_)
|
||||||
% Checks the status of posterior analysis and in particular if files need to be
|
% Checks the status of posterior analysis and in particular if files need to be
|
||||||
% created or updated; called by posterior_analysis.m
|
% created or updated; called by posterior_analysis.m
|
||||||
%
|
%
|
||||||
% Inputs:
|
% Inputs:
|
||||||
% type [string] name of the posterior moment considered
|
% type [string] name of the posterior moment considered
|
||||||
% M_ [structure] Dynare model structure
|
% M_ [structure] Dynare model structure
|
||||||
%
|
%
|
||||||
% Outputs:
|
% Outputs:
|
||||||
% info [scalar] return code
|
% info [scalar] return code
|
||||||
% info = 1; % select_posterior_draws has to be called first.
|
% info = 1; % select_posterior_draws has to be called first.
|
||||||
% info = 2; % _posterior_draws files have to be updated.
|
% info = 2; % _posterior_draws files have to be updated.
|
||||||
|
@ -17,7 +17,7 @@ function [info,description] = check_posterior_analysis_data(type,M_)
|
||||||
% info = 6; % Ok (nothing to do ;-)
|
% info = 6; % Ok (nothing to do ;-)
|
||||||
% description [string] Message corresponding to info
|
% description [string] Message corresponding to info
|
||||||
|
|
||||||
% Copyright (C) 2008-2015 Dynare Team
|
% Copyright (C) 2008-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -99,7 +99,7 @@ else
|
||||||
number_of_the_last_post_data_file = length(pdfinfo);
|
number_of_the_last_post_data_file = length(pdfinfo);
|
||||||
name_of_the_last_post_data_file = ...
|
name_of_the_last_post_data_file = ...
|
||||||
[ pwd filesep MetropolisFolder filesep ...
|
[ pwd filesep MetropolisFolder filesep ...
|
||||||
M_.fname '_' ...
|
M_.fname '_' ...
|
||||||
generic_post_data_file_name ...
|
generic_post_data_file_name ...
|
||||||
int2str(number_of_the_last_post_data_file) ...
|
int2str(number_of_the_last_post_data_file) ...
|
||||||
'.mat' ];
|
'.mat' ];
|
||||||
|
@ -108,7 +108,7 @@ else
|
||||||
info = 5; % posterior data files have to be updated.
|
info = 5; % posterior data files have to be updated.
|
||||||
if nargout>1
|
if nargout>1
|
||||||
description = 'posterior data files have to be updated.';
|
description = 'posterior data files have to be updated.';
|
||||||
end
|
end
|
||||||
else
|
else
|
||||||
info = 6; % Ok (nothing to do ;-)
|
info = 6; % Ok (nothing to do ;-)
|
||||||
if nargout>1
|
if nargout>1
|
||||||
|
|
|
@ -13,7 +13,7 @@ function [posterior_sampler_options, options_] = check_posterior_sampler_options
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2015 Dynare Team
|
% Copyright (C) 2015-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -32,32 +32,32 @@ function [posterior_sampler_options, options_] = check_posterior_sampler_options
|
||||||
|
|
||||||
|
|
||||||
init=0;
|
init=0;
|
||||||
if isempty(posterior_sampler_options),
|
if isempty(posterior_sampler_options)
|
||||||
init=1;
|
init=1;
|
||||||
end
|
end
|
||||||
|
|
||||||
if init,
|
if init
|
||||||
% set default options and user defined options
|
% set default options and user defined options
|
||||||
posterior_sampler_options.posterior_sampling_method = options_.posterior_sampler_options.posterior_sampling_method;
|
posterior_sampler_options.posterior_sampling_method = options_.posterior_sampler_options.posterior_sampling_method;
|
||||||
posterior_sampler_options.bounds = bounds;
|
posterior_sampler_options.bounds = bounds;
|
||||||
|
|
||||||
switch posterior_sampler_options.posterior_sampling_method
|
switch posterior_sampler_options.posterior_sampling_method
|
||||||
|
|
||||||
case 'random_walk_metropolis_hastings'
|
case 'random_walk_metropolis_hastings'
|
||||||
posterior_sampler_options.parallel_bar_refresh_rate=50;
|
posterior_sampler_options.parallel_bar_refresh_rate=50;
|
||||||
posterior_sampler_options.serial_bar_refresh_rate=3;
|
posterior_sampler_options.serial_bar_refresh_rate=3;
|
||||||
posterior_sampler_options.parallel_bar_title='RWMH';
|
posterior_sampler_options.parallel_bar_title='RWMH';
|
||||||
posterior_sampler_options.serial_bar_title='RW Metropolis-Hastings';
|
posterior_sampler_options.serial_bar_title='RW Metropolis-Hastings';
|
||||||
|
|
||||||
% default options
|
% default options
|
||||||
posterior_sampler_options = add_fields_(posterior_sampler_options,options_.posterior_sampler_options.rwmh);
|
posterior_sampler_options = add_fields_(posterior_sampler_options,options_.posterior_sampler_options.rwmh);
|
||||||
|
|
||||||
% user defined options
|
% user defined options
|
||||||
if ~isempty(options_.posterior_sampler_options.sampling_opt)
|
if ~isempty(options_.posterior_sampler_options.sampling_opt)
|
||||||
options_list = read_key_value_string(options_.posterior_sampler_options.sampling_opt);
|
options_list = read_key_value_string(options_.posterior_sampler_options.sampling_opt);
|
||||||
for i=1:rows(options_list)
|
for i=1:rows(options_list)
|
||||||
switch options_list{i,1}
|
switch options_list{i,1}
|
||||||
|
|
||||||
case 'proposal_distribution'
|
case 'proposal_distribution'
|
||||||
if ~(strcmpi(options_list{i,2}, 'rand_multivariate_student') || ...
|
if ~(strcmpi(options_list{i,2}, 'rand_multivariate_student') || ...
|
||||||
strcmpi(options_list{i,2}, 'rand_multivariate_normal'))
|
strcmpi(options_list{i,2}, 'rand_multivariate_normal'))
|
||||||
|
@ -66,15 +66,15 @@ if init,
|
||||||
else
|
else
|
||||||
posterior_sampler_options.proposal_distribution=options_list{i,2};
|
posterior_sampler_options.proposal_distribution=options_list{i,2};
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
||||||
case 'student_degrees_of_freedom'
|
case 'student_degrees_of_freedom'
|
||||||
if options_list{i,2} <= 0
|
if options_list{i,2} <= 0
|
||||||
error('initial_estimation_checks:: the student_degrees_of_freedom takes a positive integer argument');
|
error('initial_estimation_checks:: the student_degrees_of_freedom takes a positive integer argument');
|
||||||
else
|
else
|
||||||
posterior_sampler_options.student_degrees_of_freedom=options_list{i,2};
|
posterior_sampler_options.student_degrees_of_freedom=options_list{i,2};
|
||||||
end
|
end
|
||||||
|
|
||||||
case 'use_mh_covariance_matrix'
|
case 'use_mh_covariance_matrix'
|
||||||
% indicates to use the covariance matrix from previous iterations to
|
% indicates to use the covariance matrix from previous iterations to
|
||||||
% define the covariance of the proposal distribution
|
% define the covariance of the proposal distribution
|
||||||
|
@ -101,23 +101,23 @@ if init,
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
case 'tailored_random_block_metropolis_hastings'
|
case 'tailored_random_block_metropolis_hastings'
|
||||||
posterior_sampler_options.parallel_bar_refresh_rate=5;
|
posterior_sampler_options.parallel_bar_refresh_rate=5;
|
||||||
posterior_sampler_options.serial_bar_refresh_rate=1;
|
posterior_sampler_options.serial_bar_refresh_rate=1;
|
||||||
posterior_sampler_options.parallel_bar_title='TaRB-MH';
|
posterior_sampler_options.parallel_bar_title='TaRB-MH';
|
||||||
posterior_sampler_options.serial_bar_title='TaRB Metropolis-Hastings';
|
posterior_sampler_options.serial_bar_title='TaRB Metropolis-Hastings';
|
||||||
|
|
||||||
% default options
|
% default options
|
||||||
posterior_sampler_options = add_fields_(posterior_sampler_options,options_.posterior_sampler_options.tarb);
|
posterior_sampler_options = add_fields_(posterior_sampler_options,options_.posterior_sampler_options.tarb);
|
||||||
|
|
||||||
% user defined options
|
% user defined options
|
||||||
if ~isempty(options_.posterior_sampler_options.sampling_opt)
|
if ~isempty(options_.posterior_sampler_options.sampling_opt)
|
||||||
options_list = read_key_value_string(options_.posterior_sampler_options.sampling_opt);
|
options_list = read_key_value_string(options_.posterior_sampler_options.sampling_opt);
|
||||||
for i=1:rows(options_list)
|
for i=1:rows(options_list)
|
||||||
|
|
||||||
switch options_list{i,1}
|
switch options_list{i,1}
|
||||||
|
|
||||||
case 'proposal_distribution'
|
case 'proposal_distribution'
|
||||||
if ~(strcmpi(options_list{i,2}, 'rand_multivariate_student') || ...
|
if ~(strcmpi(options_list{i,2}, 'rand_multivariate_student') || ...
|
||||||
strcmpi(options_list{i,2}, 'rand_multivariate_normal'))
|
strcmpi(options_list{i,2}, 'rand_multivariate_normal'))
|
||||||
|
@ -126,18 +126,18 @@ if init,
|
||||||
else
|
else
|
||||||
posterior_sampler_options.proposal_distribution=options_list{i,2};
|
posterior_sampler_options.proposal_distribution=options_list{i,2};
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
||||||
case 'student_degrees_of_freedom'
|
case 'student_degrees_of_freedom'
|
||||||
if options_list{i,2} <= 0
|
if options_list{i,2} <= 0
|
||||||
error('initial_estimation_checks:: the student_degrees_of_freedom takes a positive integer argument');
|
error('initial_estimation_checks:: the student_degrees_of_freedom takes a positive integer argument');
|
||||||
else
|
else
|
||||||
posterior_sampler_options.student_degrees_of_freedom=options_list{i,2};
|
posterior_sampler_options.student_degrees_of_freedom=options_list{i,2};
|
||||||
end
|
end
|
||||||
|
|
||||||
case 'mode_compute'
|
case 'mode_compute'
|
||||||
posterior_sampler_options.mode_compute=options_list{i,2};
|
posterior_sampler_options.mode_compute=options_list{i,2};
|
||||||
|
|
||||||
case 'optim'
|
case 'optim'
|
||||||
posterior_sampler_options.optim_opt=options_list{i,2};
|
posterior_sampler_options.optim_opt=options_list{i,2};
|
||||||
|
|
||||||
|
@ -162,31 +162,31 @@ if init,
|
||||||
end
|
end
|
||||||
case 'save_tmp_file'
|
case 'save_tmp_file'
|
||||||
posterior_sampler_options.save_tmp_file = options_list{i,2};
|
posterior_sampler_options.save_tmp_file = options_list{i,2};
|
||||||
|
|
||||||
otherwise
|
otherwise
|
||||||
warning(['tarb_sampler: Unknown option (' options_list{i,1} ')!'])
|
warning(['tarb_sampler: Unknown option (' options_list{i,1} ')!'])
|
||||||
|
|
||||||
end
|
end
|
||||||
|
|
||||||
end
|
end
|
||||||
|
|
||||||
end
|
end
|
||||||
|
|
||||||
case 'independent_metropolis_hastings'
|
case 'independent_metropolis_hastings'
|
||||||
posterior_sampler_options.parallel_bar_refresh_rate=50;
|
posterior_sampler_options.parallel_bar_refresh_rate=50;
|
||||||
posterior_sampler_options.serial_bar_refresh_rate=3;
|
posterior_sampler_options.serial_bar_refresh_rate=3;
|
||||||
posterior_sampler_options.parallel_bar_title='IMH';
|
posterior_sampler_options.parallel_bar_title='IMH';
|
||||||
posterior_sampler_options.serial_bar_title='Ind. Metropolis-Hastings';
|
posterior_sampler_options.serial_bar_title='Ind. Metropolis-Hastings';
|
||||||
|
|
||||||
% default options
|
% default options
|
||||||
posterior_sampler_options = add_fields_(posterior_sampler_options,options_.posterior_sampler_options.imh);
|
posterior_sampler_options = add_fields_(posterior_sampler_options,options_.posterior_sampler_options.imh);
|
||||||
|
|
||||||
% user defined options
|
% user defined options
|
||||||
if ~isempty(options_.posterior_sampler_options.sampling_opt)
|
if ~isempty(options_.posterior_sampler_options.sampling_opt)
|
||||||
options_list = read_key_value_string(options_.posterior_sampler_options.sampling_opt);
|
options_list = read_key_value_string(options_.posterior_sampler_options.sampling_opt);
|
||||||
for i=1:rows(options_list)
|
for i=1:rows(options_list)
|
||||||
switch options_list{i,1}
|
switch options_list{i,1}
|
||||||
|
|
||||||
case 'proposal_distribution'
|
case 'proposal_distribution'
|
||||||
if ~(strcmpi(options_list{i,2}, 'rand_multivariate_student') || ...
|
if ~(strcmpi(options_list{i,2}, 'rand_multivariate_student') || ...
|
||||||
strcmpi(options_list{i,2}, 'rand_multivariate_normal'))
|
strcmpi(options_list{i,2}, 'rand_multivariate_normal'))
|
||||||
|
@ -195,22 +195,22 @@ if init,
|
||||||
else
|
else
|
||||||
posterior_sampler_options.proposal_distribution=options_list{i,2};
|
posterior_sampler_options.proposal_distribution=options_list{i,2};
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
||||||
case 'student_degrees_of_freedom'
|
case 'student_degrees_of_freedom'
|
||||||
if options_list{i,2} <= 0
|
if options_list{i,2} <= 0
|
||||||
error('initial_estimation_checks:: the student_degrees_of_freedom takes a positive integer argument');
|
error('initial_estimation_checks:: the student_degrees_of_freedom takes a positive integer argument');
|
||||||
else
|
else
|
||||||
posterior_sampler_options.student_degrees_of_freedom=options_list{i,2};
|
posterior_sampler_options.student_degrees_of_freedom=options_list{i,2};
|
||||||
end
|
end
|
||||||
|
|
||||||
case 'use_mh_covariance_matrix'
|
case 'use_mh_covariance_matrix'
|
||||||
% indicates to use the covariance matrix from previous iterations to
|
% indicates to use the covariance matrix from previous iterations to
|
||||||
% define the covariance of the proposal distribution
|
% define the covariance of the proposal distribution
|
||||||
% default = 0
|
% default = 0
|
||||||
posterior_sampler_options.use_mh_covariance_matrix = options_list{i,2};
|
posterior_sampler_options.use_mh_covariance_matrix = options_list{i,2};
|
||||||
options_.use_mh_covariance_matrix = options_list{i,2};
|
options_.use_mh_covariance_matrix = options_list{i,2};
|
||||||
|
|
||||||
case 'save_tmp_file'
|
case 'save_tmp_file'
|
||||||
posterior_sampler_options.save_tmp_file = options_list{i,2};
|
posterior_sampler_options.save_tmp_file = options_list{i,2};
|
||||||
|
|
||||||
|
@ -219,17 +219,17 @@ if init,
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
||||||
case 'slice'
|
case 'slice'
|
||||||
posterior_sampler_options.parallel_bar_refresh_rate=1;
|
posterior_sampler_options.parallel_bar_refresh_rate=1;
|
||||||
posterior_sampler_options.serial_bar_refresh_rate=1;
|
posterior_sampler_options.serial_bar_refresh_rate=1;
|
||||||
posterior_sampler_options.parallel_bar_title='SLICE';
|
posterior_sampler_options.parallel_bar_title='SLICE';
|
||||||
posterior_sampler_options.serial_bar_title='SLICE';
|
posterior_sampler_options.serial_bar_title='SLICE';
|
||||||
|
|
||||||
% default options
|
% default options
|
||||||
posterior_sampler_options = add_fields_(posterior_sampler_options,options_.posterior_sampler_options.slice);
|
posterior_sampler_options = add_fields_(posterior_sampler_options,options_.posterior_sampler_options.slice);
|
||||||
|
|
||||||
% user defined options
|
% user defined options
|
||||||
if ~isempty(options_.posterior_sampler_options.sampling_opt)
|
if ~isempty(options_.posterior_sampler_options.sampling_opt)
|
||||||
options_list = read_key_value_string(options_.posterior_sampler_options.sampling_opt);
|
options_list = read_key_value_string(options_.posterior_sampler_options.sampling_opt);
|
||||||
|
@ -242,7 +242,7 @@ if init,
|
||||||
% <use_mh_covariance_matrix> or <slice_initialize_with_mode>
|
% <use_mh_covariance_matrix> or <slice_initialize_with_mode>
|
||||||
% default = 0
|
% default = 0
|
||||||
posterior_sampler_options.rotated = options_list{i,2};
|
posterior_sampler_options.rotated = options_list{i,2};
|
||||||
|
|
||||||
case 'mode'
|
case 'mode'
|
||||||
% for multimodal posteriors, provide the list of modes as a
|
% for multimodal posteriors, provide the list of modes as a
|
||||||
% matrix, ordered by column, i.e. [x1 x2 x3] for three
|
% matrix, ordered by column, i.e. [x1 x2 x3] for three
|
||||||
|
@ -253,19 +253,19 @@ if init,
|
||||||
% This will automatically trigger <rotated>
|
% This will automatically trigger <rotated>
|
||||||
% default = []
|
% default = []
|
||||||
tmp_mode = options_list{i,2};
|
tmp_mode = options_list{i,2};
|
||||||
for j=1:size(tmp_mode,2),
|
for j=1:size(tmp_mode,2)
|
||||||
posterior_sampler_options.mode(j).m = tmp_mode(:,j);
|
posterior_sampler_options.mode(j).m = tmp_mode(:,j);
|
||||||
end
|
end
|
||||||
|
|
||||||
case 'mode_files'
|
case 'mode_files'
|
||||||
% for multimodal posteriors provide the name of
|
% for multimodal posteriors provide the name of
|
||||||
% a file containing a variable array xparams = [nparam * nmodes]
|
% a file containing a variable array xparams = [nparam * nmodes]
|
||||||
% one column per mode. With this info, the code will automatically
|
% one column per mode. With this info, the code will automatically
|
||||||
% set the <mode> option.
|
% set the <mode> option.
|
||||||
% This will automatically trigger <rotated>
|
% This will automatically trigger <rotated>
|
||||||
% default = []
|
% default = []
|
||||||
posterior_sampler_options.mode_files = options_list{i,2};
|
posterior_sampler_options.mode_files = options_list{i,2};
|
||||||
|
|
||||||
case 'slice_initialize_with_mode'
|
case 'slice_initialize_with_mode'
|
||||||
% the default for slice is to set mode_compute = 0 in the
|
% the default for slice is to set mode_compute = 0 in the
|
||||||
% preprocessor and start the chain(s) from a random location in the prior.
|
% preprocessor and start the chain(s) from a random location in the prior.
|
||||||
|
@ -275,7 +275,7 @@ if init,
|
||||||
% iterations.
|
% iterations.
|
||||||
% default = 0
|
% default = 0
|
||||||
posterior_sampler_options.slice_initialize_with_mode = options_list{i,2};
|
posterior_sampler_options.slice_initialize_with_mode = options_list{i,2};
|
||||||
|
|
||||||
case 'initial_step_size'
|
case 'initial_step_size'
|
||||||
% sets the initial size of the interval in the STEPPING-OUT PROCEDURE
|
% sets the initial size of the interval in the STEPPING-OUT PROCEDURE
|
||||||
% the initial_step_size must be a real number in [0, 1],
|
% the initial_step_size must be a real number in [0, 1],
|
||||||
|
@ -298,18 +298,18 @@ if init,
|
||||||
|
|
||||||
case 'save_tmp_file'
|
case 'save_tmp_file'
|
||||||
posterior_sampler_options.save_tmp_file = options_list{i,2};
|
posterior_sampler_options.save_tmp_file = options_list{i,2};
|
||||||
|
|
||||||
otherwise
|
otherwise
|
||||||
warning(['slice_sampler: Unknown option (' options_list{i,1} ')!'])
|
warning(['slice_sampler: Unknown option (' options_list{i,1} ')!'])
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
% slice posterior sampler does not require mode or hessian to run
|
% slice posterior sampler does not require mode or hessian to run
|
||||||
% needs to be set to 1 to skip parts in dynare_estimation_1.m
|
% needs to be set to 1 to skip parts in dynare_estimation_1.m
|
||||||
% requiring posterior maximization/calibrated smoother before MCMC
|
% requiring posterior maximization/calibrated smoother before MCMC
|
||||||
options_.mh_posterior_mode_estimation=1;
|
options_.mh_posterior_mode_estimation=1;
|
||||||
|
|
||||||
if ~ posterior_sampler_options.slice_initialize_with_mode
|
if ~ posterior_sampler_options.slice_initialize_with_mode
|
||||||
% by default, slice sampler should trigger
|
% by default, slice sampler should trigger
|
||||||
% mode_compute=0 and
|
% mode_compute=0 and
|
||||||
|
@ -327,52 +327,52 @@ if init,
|
||||||
options_.mh_posterior_mode_estimation=0;
|
options_.mh_posterior_mode_estimation=0;
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
if any(isinf(bounds.lb)) || any(isinf(bounds.ub)),
|
if any(isinf(bounds.lb)) || any(isinf(bounds.ub))
|
||||||
skipline()
|
skipline()
|
||||||
disp('some priors are unbounded and prior_trunc is set to zero')
|
disp('some priors are unbounded and prior_trunc is set to zero')
|
||||||
error('The option "slice" is inconsistent with prior_trunc=0.')
|
error('The option "slice" is inconsistent with prior_trunc=0.')
|
||||||
end
|
end
|
||||||
|
|
||||||
% moreover slice must be associated to:
|
% moreover slice must be associated to:
|
||||||
% options_.mh_posterior_mode_estimation = 0;
|
% options_.mh_posterior_mode_estimation = 0;
|
||||||
% this is done below, but perhaps preprocessing should do this?
|
% this is done below, but perhaps preprocessing should do this?
|
||||||
|
|
||||||
if ~isempty(posterior_sampler_options.mode)
|
if ~isempty(posterior_sampler_options.mode)
|
||||||
% multimodal case
|
% multimodal case
|
||||||
posterior_sampler_options.rotated = 1;
|
posterior_sampler_options.rotated = 1;
|
||||||
posterior_sampler_options.WR=[];
|
posterior_sampler_options.WR=[];
|
||||||
end
|
end
|
||||||
% posterior_sampler_options = set_default_option(posterior_sampler_options,'mode_files',[]);
|
% posterior_sampler_options = set_default_option(posterior_sampler_options,'mode_files',[]);
|
||||||
|
|
||||||
|
|
||||||
posterior_sampler_options.W1=posterior_sampler_options.initial_step_size*(bounds.ub-bounds.lb);
|
posterior_sampler_options.W1=posterior_sampler_options.initial_step_size*(bounds.ub-bounds.lb);
|
||||||
if options_.load_mh_file,
|
if options_.load_mh_file
|
||||||
posterior_sampler_options.slice_initialize_with_mode = 0;
|
posterior_sampler_options.slice_initialize_with_mode = 0;
|
||||||
else
|
else
|
||||||
if ~posterior_sampler_options.slice_initialize_with_mode,
|
if ~posterior_sampler_options.slice_initialize_with_mode
|
||||||
posterior_sampler_options.invhess=[];
|
posterior_sampler_options.invhess=[];
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
if ~isempty(posterior_sampler_options.mode_files), % multimodal case
|
if ~isempty(posterior_sampler_options.mode_files) % multimodal case
|
||||||
modes = posterior_sampler_options.mode_files; % these can be also mean files from previous parallel slice chains
|
modes = posterior_sampler_options.mode_files; % these can be also mean files from previous parallel slice chains
|
||||||
load(modes, 'xparams')
|
load(modes, 'xparams')
|
||||||
if size(xparams,2)<2,
|
if size(xparams,2)<2
|
||||||
error(['check_posterior_sampler_options:: Variable xparams loaded in file <' modes '> has size [' int2str(size(xparams,1)) 'x' int2str(size(xparams,2)) ']: it must contain at least two columns, to allow multi-modal sampling.'])
|
error(['check_posterior_sampler_options:: Variable xparams loaded in file <' modes '> has size [' int2str(size(xparams,1)) 'x' int2str(size(xparams,2)) ']: it must contain at least two columns, to allow multi-modal sampling.'])
|
||||||
end
|
end
|
||||||
for j=1:size(xparams,2),
|
for j=1:size(xparams,2)
|
||||||
mode(j).m=xparams(:,j);
|
mode(j).m=xparams(:,j);
|
||||||
end
|
end
|
||||||
posterior_sampler_options.mode = mode;
|
posterior_sampler_options.mode = mode;
|
||||||
posterior_sampler_options.rotated = 1;
|
posterior_sampler_options.rotated = 1;
|
||||||
posterior_sampler_options.WR=[];
|
posterior_sampler_options.WR=[];
|
||||||
end
|
end
|
||||||
|
|
||||||
otherwise
|
otherwise
|
||||||
error('check_posterior_sampler_options:: Unknown posterior_sampling_method option %s ',posterior_sampler_options.posterior_sampling_method);
|
error('check_posterior_sampler_options:: Unknown posterior_sampling_method option %s ',posterior_sampler_options.posterior_sampling_method);
|
||||||
end
|
end
|
||||||
|
|
||||||
return
|
return
|
||||||
end
|
end
|
||||||
|
|
||||||
|
@ -386,7 +386,7 @@ if ~strcmp(posterior_sampler_options.posterior_sampling_method,'slice')
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
if options_.load_mh_file && posterior_sampler_options.use_mh_covariance_matrix,
|
if options_.load_mh_file && posterior_sampler_options.use_mh_covariance_matrix
|
||||||
[junk, invhess] = compute_mh_covariance_matrix;
|
[junk, invhess] = compute_mh_covariance_matrix;
|
||||||
posterior_sampler_options.invhess = invhess;
|
posterior_sampler_options.invhess = invhess;
|
||||||
end
|
end
|
||||||
|
@ -396,10 +396,8 @@ end
|
||||||
% check specific options for slice sampler
|
% check specific options for slice sampler
|
||||||
if strcmp(posterior_sampler_options.posterior_sampling_method,'slice')
|
if strcmp(posterior_sampler_options.posterior_sampling_method,'slice')
|
||||||
invhess = posterior_sampler_options.invhess;
|
invhess = posterior_sampler_options.invhess;
|
||||||
|
if posterior_sampler_options.rotated
|
||||||
if posterior_sampler_options.rotated,
|
if isempty(posterior_sampler_options.mode_files) && isempty(posterior_sampler_options.mode) % rotated unimodal
|
||||||
if isempty(posterior_sampler_options.mode_files) && isempty(posterior_sampler_options.mode), % rotated unimodal
|
|
||||||
|
|
||||||
if ~options_.cova_compute && ~(options_.load_mh_file && posterior_sampler_options.use_mh_covariance_matrix)
|
if ~options_.cova_compute && ~(options_.load_mh_file && posterior_sampler_options.use_mh_covariance_matrix)
|
||||||
skipline()
|
skipline()
|
||||||
disp('check_posterior_sampler_options:: I cannot start rotated slice sampler because')
|
disp('check_posterior_sampler_options:: I cannot start rotated slice sampler because')
|
||||||
|
@ -419,19 +417,13 @@ if strcmp(posterior_sampler_options.posterior_sampling_method,'slice')
|
||||||
posterior_sampler_options.WR=sqrt(diag(D))*3;
|
posterior_sampler_options.WR=sqrt(diag(D))*3;
|
||||||
end
|
end
|
||||||
else
|
else
|
||||||
if ~options_.load_mh_file && ~posterior_sampler_options.slice_initialize_with_mode,
|
if ~options_.load_mh_file && ~posterior_sampler_options.slice_initialize_with_mode
|
||||||
posterior_sampler_options.invhess=[];
|
posterior_sampler_options.invhess=[];
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
% needs to be re-set to zero otherwise posterior analysis is filtered
|
% needs to be re-set to zero otherwise posterior analysis is filtered
|
||||||
% out in dynare_estimation_1.m
|
% out in dynare_estimation_1.m
|
||||||
options_.mh_posterior_mode_estimation = 0;
|
options_.mh_posterior_mode_estimation = 0;
|
||||||
|
|
||||||
|
|
||||||
else
|
|
||||||
|
|
||||||
|
|
||||||
end
|
end
|
||||||
|
|
||||||
return
|
return
|
||||||
|
@ -442,6 +434,3 @@ fnam = fieldnames(sampler_options);
|
||||||
for j=1:length(fnam)
|
for j=1:length(fnam)
|
||||||
posterior_sampler_options.(fnam{j}) = sampler_options.(fnam{j});
|
posterior_sampler_options.(fnam{j}) = sampler_options.(fnam{j});
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
|
@ -1,13 +1,13 @@
|
||||||
function [info,description] = check_prior_analysis_data(type,M_)
|
function [info,description] = check_prior_analysis_data(type,M_)
|
||||||
% function [info,description] = check_prior_analysis_data(type,M_)
|
% function [info,description] = check_prior_analysis_data(type,M_)
|
||||||
% Checks the status of prior analysis and in particular if files need to be
|
% Checks the status of prior analysis and in particular if files need to be
|
||||||
% created or updated; called by prior_analysis.m
|
% created or updated; called by prior_analysis.m
|
||||||
%
|
%
|
||||||
% Inputs:
|
% Inputs:
|
||||||
% type [string] name of the posterior moment considered
|
% type [string] name of the posterior moment considered
|
||||||
% M_ [structure] Dynare model structure
|
% M_ [structure] Dynare model structure
|
||||||
%
|
%
|
||||||
% Outputs:
|
% Outputs:
|
||||||
% info [scalar] return code
|
% info [scalar] return code
|
||||||
% info = 1; % prior_sampler has to be called first.
|
% info = 1; % prior_sampler has to be called first.
|
||||||
% info = 2; % _prior_draws files have to be updated.
|
% info = 2; % _prior_draws files have to be updated.
|
||||||
|
@ -18,7 +18,7 @@ function [info,description] = check_prior_analysis_data(type,M_)
|
||||||
% description [string] Message corresponding to info
|
% description [string] Message corresponding to info
|
||||||
|
|
||||||
|
|
||||||
% Copyright (C) 2009-2011 Dynare Team
|
% Copyright (C) 2009-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -40,7 +40,7 @@ if nargout>1
|
||||||
description = '';
|
description = '';
|
||||||
end
|
end
|
||||||
|
|
||||||
%% Get informations about prior draws files.
|
%% Get informations about prior draws files.
|
||||||
if ~exist([ M_.dname '/prior/draws'],'dir')
|
if ~exist([ M_.dname '/prior/draws'],'dir')
|
||||||
disp('check_prior_analysis_data:: Can''t find any prior draws file!')
|
disp('check_prior_analysis_data:: Can''t find any prior draws file!')
|
||||||
return
|
return
|
||||||
|
@ -65,7 +65,7 @@ else
|
||||||
end
|
end
|
||||||
return
|
return
|
||||||
else
|
else
|
||||||
info = 3; % Nothing to do!
|
info = 3; % Nothing to do!
|
||||||
if nargout>1
|
if nargout>1
|
||||||
description = 'prior draws files are up to date.';
|
description = 'prior draws files are up to date.';
|
||||||
end
|
end
|
||||||
|
@ -98,9 +98,9 @@ else
|
||||||
name_of_the_last_prior_data_file = pdinfo(end).name;
|
name_of_the_last_prior_data_file = pdinfo(end).name;
|
||||||
pdfdate = pdinfo(end).datenum;
|
pdfdate = pdinfo(end).datenum;
|
||||||
% /!\ REMARK /!\
|
% /!\ REMARK /!\
|
||||||
% The user can change the model or the value of a calibrated
|
% The user can change the model or the value of a calibrated
|
||||||
% parameter without changing the prior. In this case the (prior)
|
% parameter without changing the prior. In this case the (prior)
|
||||||
% moments should be computed. But this case cannot be detected!!!
|
% moments should be computed. But this case cannot be detected!!!
|
||||||
if pdfdate<date_of_the_last_prior_draw_file
|
if pdfdate<date_of_the_last_prior_draw_file
|
||||||
info = 5; % prior data files have to be updated.
|
info = 5; % prior data files have to be updated.
|
||||||
if nargout>1
|
if nargout>1
|
||||||
|
|
|
@ -5,12 +5,12 @@ function check_prior_bounds(xparam1,bounds,M_,estim_params_,options_,bayestopt_)
|
||||||
% -xparam1 [double] vector of parameters to be estimated (initial values)
|
% -xparam1 [double] vector of parameters to be estimated (initial values)
|
||||||
% -bounds [vector] vector containing the lower and upper
|
% -bounds [vector] vector containing the lower and upper
|
||||||
% bounds
|
% bounds
|
||||||
% -M_ [structure] characterizing the model.
|
% -M_ [structure] characterizing the model.
|
||||||
% -estim_params_ [structure] characterizing parameters to be estimated
|
% -estim_params_ [structure] characterizing parameters to be estimated
|
||||||
% -options_ [structure] characterizing the options
|
% -options_ [structure] characterizing the options
|
||||||
% -bayestopt_ [structure] characterizing priors
|
% -bayestopt_ [structure] characterizing priors
|
||||||
|
|
||||||
% Copyright (C) 2013 Dynare Team
|
% Copyright (C) 2013-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
|
|
@ -1,17 +1,17 @@
|
||||||
function [R,indef, E, P]=chol_SE(A,pivoting)
|
function [R,indef, E, P]=chol_SE(A,pivoting)
|
||||||
% [R,indef, E, P]=chol_SE(A,pivoting)
|
% [R,indef, E, P]=chol_SE(A,pivoting)
|
||||||
% Performs a (modified) Cholesky factorization of the form
|
% Performs a (modified) Cholesky factorization of the form
|
||||||
%
|
%
|
||||||
% P'*A*P + E = R'*R
|
% P'*A*P + E = R'*R
|
||||||
%
|
%
|
||||||
% As detailed in Schnabel/Eskow (1990), the factorization has 2 phases:
|
% As detailed in Schnabel/Eskow (1990), the factorization has 2 phases:
|
||||||
% Phase 1: While A can still be positive definite, pivot on the maximum diagonal element.
|
% Phase 1: While A can still be positive definite, pivot on the maximum diagonal element.
|
||||||
% Check that the standard Cholesky update would result in a positive diagonal
|
% Check that the standard Cholesky update would result in a positive diagonal
|
||||||
% at the current iteration. If so, continue with the normal Cholesky update.
|
% at the current iteration. If so, continue with the normal Cholesky update.
|
||||||
% Otherwise switch to phase 2.
|
% Otherwise switch to phase 2.
|
||||||
% If A is safely positive definite, stage 1 is never left, resulting in
|
% If A is safely positive definite, stage 1 is never left, resulting in
|
||||||
% the standard Cholesky decomposition.
|
% the standard Cholesky decomposition.
|
||||||
%
|
%
|
||||||
% Phase 2: Pivot on the minimum of the negatives of the lower Gershgorin bound
|
% Phase 2: Pivot on the minimum of the negatives of the lower Gershgorin bound
|
||||||
% estimates. To prevent negative diagonals, compute the amount to add to the
|
% estimates. To prevent negative diagonals, compute the amount to add to the
|
||||||
% pivot element and add this. Then, do the Cholesky update and update the estimates of the
|
% pivot element and add this. Then, do the Cholesky update and update the estimates of the
|
||||||
|
@ -21,39 +21,40 @@ function [R,indef, E, P]=chol_SE(A,pivoting)
|
||||||
% - During factorization, L=R' is stored in the lower triangle of the original matrix A,
|
% - During factorization, L=R' is stored in the lower triangle of the original matrix A,
|
||||||
% miminizing the memory requirements
|
% miminizing the memory requirements
|
||||||
% - Conforming with the original Schnabel/Eskow (1990) algorithm:
|
% - Conforming with the original Schnabel/Eskow (1990) algorithm:
|
||||||
% - at each iteration the updated Gershgorin bounds are estimated instead of recomputed,
|
% - at each iteration the updated Gershgorin bounds are estimated instead of recomputed,
|
||||||
% reducing the computational requirements from o(n^3) to o (n^2)
|
% reducing the computational requirements from o(n^3) to o (n^2)
|
||||||
% - For the last 2 by 2 submatrix, an eigenvalue-based decomposition is used
|
% - For the last 2 by 2 submatrix, an eigenvalue-based decomposition is used
|
||||||
% - While pivoting is not necessary, it improves the size of E, the add-on on to the diagonal. But this comes at
|
% - While pivoting is not necessary, it improves the size of E, the add-on on to the diagonal. But this comes at
|
||||||
% the cost of introduding a permutation.
|
% the cost of introduding a permutation.
|
||||||
%
|
|
||||||
%
|
%
|
||||||
% Inputs
|
|
||||||
% A [n*n] Matrix to be factorized
|
|
||||||
% pivoting [scalar] dummy whether pivoting is used
|
|
||||||
%
|
|
||||||
% Outputs
|
|
||||||
% R [n*n] originally stored in lower triangular portion of A, including the main diagonal
|
|
||||||
%
|
%
|
||||||
% E [n*1] Elements added to the diagonal of A
|
% INPUTS
|
||||||
% P [n*1] record of how the rows and columns of the matrix were permuted while
|
% - A [n*n] Matrix to be factorized
|
||||||
% performing the decomposition
|
% - pivoting [scalar] dummy whether pivoting is used
|
||||||
|
%
|
||||||
|
% OUTPUTS
|
||||||
|
% - R [n*n] originally stored in lower triangular portion of A, including the main diagonal
|
||||||
|
%
|
||||||
|
% - E [n*1] Elements added to the diagonal of A
|
||||||
|
% - P [n*1] record of how the rows and columns of the matrix were permuted while
|
||||||
|
% performing the decomposition
|
||||||
%
|
%
|
||||||
% REFERENCES:
|
% REFERENCES:
|
||||||
% This implementation is based on
|
% This implementation is based on
|
||||||
%
|
%
|
||||||
% Robert B. Schnabel and Elizabeth Eskow. 1990. "A New Modified Cholesky
|
% Robert B. Schnabel and Elizabeth Eskow. 1990. "A New Modified Cholesky
|
||||||
% Factorization," SIAM Journal of Scientific Statistical Computating,
|
% Factorization," SIAM Journal of Scientific Statistical Computating,
|
||||||
% 11, 6: 1136-58.
|
% 11, 6: 1136-58.
|
||||||
%
|
%
|
||||||
% Elizabeth Eskow and Robert B. Schnabel 1991. "Algorithm 695 - Software for a New Modified Cholesky
|
% Elizabeth Eskow and Robert B. Schnabel 1991. "Algorithm 695 - Software for a New Modified Cholesky
|
||||||
% Factorization," ACM Transactions on Mathematical Software, Vol 17, No 3: 306-312
|
% Factorization," ACM Transactions on Mathematical Software, Vol 17, No 3: 306-312
|
||||||
%
|
%
|
||||||
%
|
%
|
||||||
% Author: Johannes Pfeifer based on Eskow/Schnabel (1991)
|
% Author: Johannes Pfeifer based on Eskow/Schnabel (1991)
|
||||||
%
|
|
||||||
% Copyright (C) 2015 Johannes Pfeifer
|
% Copyright (C) 2015 Johannes Pfeifer
|
||||||
% Copyright (C) 2015 Dynare Team
|
% Copyright (C) 2015-2017 Dynare Team
|
||||||
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
% Dynare is free software: you can redistribute it and/or modify
|
% Dynare is free software: you can redistribute it and/or modify
|
||||||
|
@ -148,7 +149,7 @@ for j = 1:n-1
|
||||||
% Swap elements of the permutation vector
|
% Swap elements of the permutation vector
|
||||||
itemp = P(j);
|
itemp = P(j);
|
||||||
P(j) = P(imaxd);
|
P(j) = P(imaxd);
|
||||||
P(imaxd) = itemp;
|
P(imaxd) = itemp;
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
% check to see whether the normal Cholesky update for this
|
% check to see whether the normal Cholesky update for this
|
||||||
|
@ -183,13 +184,13 @@ for j = 1:n-1
|
||||||
g=gersh_nested(A,j,n);
|
g=gersh_nested(A,j,n);
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
% PHASE 2
|
% PHASE 2
|
||||||
if ~phase1
|
if ~phase1
|
||||||
if j ~= n-1
|
if j ~= n-1
|
||||||
if pivoting
|
if pivoting
|
||||||
% Find the minimum negative Gershgorin bound
|
% Find the minimum negative Gershgorin bound
|
||||||
[tmp,iming] = min(g(j:n));
|
[tmp,iming] = min(g(j:n));
|
||||||
iming=iming+j-1;
|
iming=iming+j-1;
|
||||||
% Pivot to the top the row and column with the
|
% Pivot to the top the row and column with the
|
||||||
% minimum negative Gershgorin bound
|
% minimum negative Gershgorin bound
|
||||||
|
@ -229,9 +230,9 @@ for j = 1:n-1
|
||||||
end
|
end
|
||||||
% Calculate delta and add to the diagonal. delta=max{0,-A(j,j) + max{normj,taugam},delta_previous}
|
% Calculate delta and add to the diagonal. delta=max{0,-A(j,j) + max{normj,taugam},delta_previous}
|
||||||
% where normj=sum of |A(i,j)|,for i=1,n, delta_previous is the delta computed at the previous iter and taugam is tau1*gamma.
|
% where normj=sum of |A(i,j)|,for i=1,n, delta_previous is the delta computed at the previous iter and taugam is tau1*gamma.
|
||||||
|
|
||||||
normj=sum(abs(A(j+1:n,j)));
|
normj=sum(abs(A(j+1:n,j)));
|
||||||
|
|
||||||
delta = max([0;delta;-A(j,j)+normj;-A(j,j)+taugam]); % get adjustment based on formula on bottom of p. 309 of Eskow/Schnabel (1991)
|
delta = max([0;delta;-A(j,j)+normj;-A(j,j)+taugam]); % get adjustment based on formula on bottom of p. 309 of Eskow/Schnabel (1991)
|
||||||
|
|
||||||
E(j) = delta;
|
E(j) = delta;
|
||||||
|
@ -259,9 +260,9 @@ for j = 1:n-1
|
||||||
% Find eigenvalues of final 2 by 2 submatrix
|
% Find eigenvalues of final 2 by 2 submatrix
|
||||||
% Find delta such that:
|
% Find delta such that:
|
||||||
% 1. the l2 condition number of the final 2X2 submatrix + delta*I <= tau2
|
% 1. the l2 condition number of the final 2X2 submatrix + delta*I <= tau2
|
||||||
% 2. delta >= previous delta,
|
% 2. delta >= previous delta,
|
||||||
% 3. min(eigvals) + delta >= tau2 * gamma, where min(eigvals) is the smallest eigenvalue of the final 2X2 submatrix
|
% 3. min(eigvals) + delta >= tau2 * gamma, where min(eigvals) is the smallest eigenvalue of the final 2X2 submatrix
|
||||||
|
|
||||||
A(n-1,n)=A(n,n-1); %set value above diagonal for computation of eigenvalues
|
A(n-1,n)=A(n,n-1); %set value above diagonal for computation of eigenvalues
|
||||||
eigvals = eig(A(n-1:n,n-1:n));
|
eigvals = eig(A(n-1:n,n-1:n));
|
||||||
delta = max([ 0 ; delta ; -min(eigvals)+tau2*max((max(eigvals)-min(eigvals))/(1-tau1),gammma) ]); %Formula 5.3.2 of Schnabel/Eskow (1990)
|
delta = max([ 0 ; delta ; -min(eigvals)+tau2*max((max(eigvals)-min(eigvals))/(1-tau1),gammma) ]); %Formula 5.3.2 of Schnabel/Eskow (1990)
|
||||||
|
@ -272,7 +273,7 @@ for j = 1:n-1
|
||||||
E(n-1) = delta;
|
E(n-1) = delta;
|
||||||
E(n) = delta;
|
E(n) = delta;
|
||||||
end
|
end
|
||||||
|
|
||||||
% Final update
|
% Final update
|
||||||
A(n-1,n-1) = sqrt(A(n-1,n-1));
|
A(n-1,n-1) = sqrt(A(n-1,n-1));
|
||||||
A(n,n-1) = A(n,n-1)/A(n-1,n-1);
|
A(n,n-1) = A(n,n-1)/A(n-1,n-1);
|
||||||
|
@ -293,17 +294,16 @@ function g=gersh_nested(A,j,n)
|
||||||
|
|
||||||
g=zeros(n,1);
|
g=zeros(n,1);
|
||||||
for ii = j:n
|
for ii = j:n
|
||||||
if ii == 1;
|
if ii == 1
|
||||||
sum_up_to_i = 0;
|
sum_up_to_i = 0;
|
||||||
else
|
else
|
||||||
sum_up_to_i = sum(abs(A(ii,j:(ii-1))));
|
sum_up_to_i = sum(abs(A(ii,j:(ii-1))));
|
||||||
end;
|
end
|
||||||
if ii == n;
|
if ii == n
|
||||||
sum_after_i = 0;
|
sum_after_i = 0;
|
||||||
else
|
else
|
||||||
sum_after_i = sum(abs(A((ii+1):n,ii)));
|
sum_after_i = sum(abs(A((ii+1):n,ii)));
|
||||||
end;
|
end
|
||||||
g(ii) = sum_up_to_i + sum_after_i- A(ii,ii);
|
g(ii) = sum_up_to_i + sum_after_i- A(ii,ii);
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
|
@ -2,7 +2,7 @@ function clear_persistent_variables(folder, writelistofroutinestobecleared)
|
||||||
|
|
||||||
% Clear all the functions with persistent variables in directory folder (and subdirectories).
|
% Clear all the functions with persistent variables in directory folder (and subdirectories).
|
||||||
|
|
||||||
% Copyright (C) 2015 Dynare Team
|
% Copyright (C) 2015-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -58,5 +58,5 @@ if writelistofroutinestobecleared
|
||||||
return
|
return
|
||||||
end
|
end
|
||||||
|
|
||||||
list_of_functions_to_be_cleared;
|
list_of_functions_to_be_cleared;
|
||||||
clear(list_of_functions{:});
|
clear(list_of_functions{:});
|
|
@ -11,7 +11,7 @@ function varargout = prior(varargin)
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2015-2016 Dynare Team
|
% Copyright (C) 2015-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -42,7 +42,7 @@ if isempty(varargin) || ( isequal(length(varargin), 1) && isequal(varargin{1},'h
|
||||||
return
|
return
|
||||||
end
|
end
|
||||||
|
|
||||||
global options_ M_ estim_params_ bayestopt_ oo_
|
global options_ M_ estim_params_ bayestopt_ oo_
|
||||||
|
|
||||||
donesomething = false;
|
donesomething = false;
|
||||||
|
|
||||||
|
|
|
@ -1,14 +1,14 @@
|
||||||
function collect_latex_files
|
function collect_latex_files
|
||||||
% function collect_LaTeX_Files;
|
% function collect_LaTeX_Files;
|
||||||
% Creates TeX-File embedding all eps-loaders created for current mod-file
|
% Creates TeX-File embedding all eps-loaders created for current mod-file
|
||||||
%
|
%
|
||||||
% Inputs: none
|
% Inputs: none
|
||||||
%
|
%
|
||||||
% Notes:
|
% Notes:
|
||||||
% - The packages loaded enable pdflatex to run
|
% - The packages loaded enable pdflatex to run
|
||||||
% - The _dynamic and _static TeX-model files are not included as they are standalone TeX-files
|
% - The _dynamic and _static TeX-model files are not included as they are standalone TeX-files
|
||||||
|
|
||||||
% Copyright (C) 2015-16 Dynare Team
|
% Copyright (C) 2015-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -49,7 +49,7 @@ for ii=1:length(TeX_Files)
|
||||||
~strcmp(TeX_Files(ii).name,[M_.fname,'_static.tex']) && ...
|
~strcmp(TeX_Files(ii).name,[M_.fname,'_static.tex']) && ...
|
||||||
~strcmp(TeX_Files(ii).name,[M_.fname,'_original.tex']) && ...
|
~strcmp(TeX_Files(ii).name,[M_.fname,'_original.tex']) && ...
|
||||||
~strcmp(TeX_Files(ii).name,[M_.fname,'_TeX_binder.tex'])
|
~strcmp(TeX_Files(ii).name,[M_.fname,'_TeX_binder.tex'])
|
||||||
fprintf(fid,'%s \n',['\include{',f_name,'}']);
|
fprintf(fid,'%s \n',['\include{',f_name,'}']);
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
|
@ -58,7 +58,7 @@ TeX_Files=dir([M_.dname filesep 'Output' filesep M_.fname '*.tex']);
|
||||||
for ii=1:length(TeX_Files)
|
for ii=1:length(TeX_Files)
|
||||||
[pathstr,f_name,ext] = fileparts(TeX_Files(ii).name);
|
[pathstr,f_name,ext] = fileparts(TeX_Files(ii).name);
|
||||||
if ~strcmp(TeX_Files(ii).name,f_name_binder)
|
if ~strcmp(TeX_Files(ii).name,f_name_binder)
|
||||||
fprintf(fid,'%s \n',['\include{', M_.dname '/Output' '/',f_name,'}']);
|
fprintf(fid,'%s \n',['\include{', M_.dname '/Output' '/',f_name,'}']);
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
|
@ -67,7 +67,7 @@ TeX_Files=dir([M_.dname filesep 'graphs' filesep M_.fname '*.tex']);
|
||||||
for ii=1:length(TeX_Files)
|
for ii=1:length(TeX_Files)
|
||||||
[pathstr,f_name,ext] = fileparts(TeX_Files(ii).name);
|
[pathstr,f_name,ext] = fileparts(TeX_Files(ii).name);
|
||||||
if ~strcmp(TeX_Files(ii).name,f_name_binder)
|
if ~strcmp(TeX_Files(ii).name,f_name_binder)
|
||||||
fprintf(fid,'%s \n',['\include{', M_.dname '/graphs' '/',f_name,'}']);
|
fprintf(fid,'%s \n',['\include{', M_.dname '/graphs' '/',f_name,'}']);
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
|
@ -76,7 +76,7 @@ TeX_Files=dir([M_.dname filesep 'identification' filesep M_.fname '*.tex']);
|
||||||
for ii=1:length(TeX_Files)
|
for ii=1:length(TeX_Files)
|
||||||
[pathstr,f_name,ext] = fileparts(TeX_Files(ii).name);
|
[pathstr,f_name,ext] = fileparts(TeX_Files(ii).name);
|
||||||
if ~strcmp(TeX_Files(ii).name,f_name_binder)
|
if ~strcmp(TeX_Files(ii).name,f_name_binder)
|
||||||
fprintf(fid,'%s \n',['\include{', M_.dname '/identification' '/',f_name,'}']);
|
fprintf(fid,'%s \n',['\include{', M_.dname '/identification' '/',f_name,'}']);
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
|
@ -86,7 +86,7 @@ TeX_Files=dir([M_.dname filesep 'identification' filesep 'Output' filesep M_.fna
|
||||||
for ii=1:length(TeX_Files)
|
for ii=1:length(TeX_Files)
|
||||||
[pathstr,f_name,ext] = fileparts(TeX_Files(ii).name);
|
[pathstr,f_name,ext] = fileparts(TeX_Files(ii).name);
|
||||||
if ~strcmp(TeX_Files(ii).name,f_name_binder)
|
if ~strcmp(TeX_Files(ii).name,f_name_binder)
|
||||||
fprintf(fid,'%s \n',['\include{', M_.dname '/identification/Output' '/',f_name,'}']);
|
fprintf(fid,'%s \n',['\include{', M_.dname '/identification/Output' '/',f_name,'}']);
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
|
@ -95,7 +95,7 @@ TeX_Files=dir([M_.dname filesep 'gsa' filesep M_.fname '*.tex']);
|
||||||
for ii=1:length(TeX_Files)
|
for ii=1:length(TeX_Files)
|
||||||
[pathstr,f_name,ext] = fileparts(TeX_Files(ii).name);
|
[pathstr,f_name,ext] = fileparts(TeX_Files(ii).name);
|
||||||
if ~strcmp(TeX_Files(ii).name,f_name_binder)
|
if ~strcmp(TeX_Files(ii).name,f_name_binder)
|
||||||
fprintf(fid,'%s \n',['\include{', M_.dname '/gsa' '/',f_name,'}']);
|
fprintf(fid,'%s \n',['\include{', M_.dname '/gsa' '/',f_name,'}']);
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
|
@ -104,7 +104,7 @@ TeX_Files=dir([M_.dname filesep 'gsa' filesep 'Output' filesep M_.fname '*.tex'
|
||||||
for ii=1:length(TeX_Files)
|
for ii=1:length(TeX_Files)
|
||||||
[pathstr,f_name,ext] = fileparts(TeX_Files(ii).name);
|
[pathstr,f_name,ext] = fileparts(TeX_Files(ii).name);
|
||||||
if ~strcmp(TeX_Files(ii).name,f_name_binder)
|
if ~strcmp(TeX_Files(ii).name,f_name_binder)
|
||||||
fprintf(fid,'%s \n',['\include{', M_.dname '/gsa/Output' '/',f_name,'}']);
|
fprintf(fid,'%s \n',['\include{', M_.dname '/gsa/Output' '/',f_name,'}']);
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
|
@ -1,36 +1,36 @@
|
||||||
function [Pstar,Pinf] = compute_Pinf_Pstar(mf,T,R,Q,qz_criterium, restrict_columns)
|
function [Pstar,Pinf] = compute_Pinf_Pstar(mf,T,R,Q,qz_criterium, restrict_columns)
|
||||||
% function [Z,ST,QT,R1,Pstar,Pinf] = schur_statespace_transformation(mf,T,R,Q,qz_criterium, restrict_columns)
|
% function [Z,ST,QT,R1,Pstar,Pinf] = schur_statespace_transformation(mf,T,R,Q,qz_criterium, restrict_columns)
|
||||||
% Kitagawa transformation of state space system with a quasi-triangular
|
% Kitagawa transformation of state space system with a quasi-triangular
|
||||||
% transition matrix with unit roots at the top, but excluding zero columns of the transition matrix.
|
% transition matrix with unit roots at the top, but excluding zero columns of the transition matrix.
|
||||||
% Computation of Pstar and Pinf for Durbin and Koopman Diffuse filter
|
% Computation of Pstar and Pinf for Durbin and Koopman Diffuse filter
|
||||||
%
|
%
|
||||||
% The transformed state space is
|
% The transformed state space is
|
||||||
% y = [ss; z; x];
|
% y = [ss; z; x];
|
||||||
% s = static variables (zero columns of T)
|
% s = static variables (zero columns of T)
|
||||||
% z = unit roots
|
% z = unit roots
|
||||||
% x = stable roots
|
% x = stable roots
|
||||||
% ss = s - z = stationarized static variables
|
% ss = s - z = stationarized static variables
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% mf [integer] vector of indices of observed variables in
|
% mf [integer] vector of indices of observed variables in
|
||||||
% state vector
|
% state vector
|
||||||
% T [double] matrix of transition
|
% T [double] matrix of transition
|
||||||
% R [double] matrix of structural shock effects
|
% R [double] matrix of structural shock effects
|
||||||
% Q [double] matrix of covariance of structural shocks
|
% Q [double] matrix of covariance of structural shocks
|
||||||
% qz_criterium [double] numerical criterium for unit roots
|
% qz_criterium [double] numerical criterium for unit roots
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% Pstar [double] matrix of covariance of stationary part
|
% Pstar [double] matrix of covariance of stationary part
|
||||||
% Pinf [double] matrix of covariance initialization for
|
% Pinf [double] matrix of covariance initialization for
|
||||||
% nonstationary part
|
% nonstationary part
|
||||||
%
|
%
|
||||||
% ALGORITHM
|
% ALGORITHM
|
||||||
% Real Schur transformation of transition equation
|
% Real Schur transformation of transition equation
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% None
|
% None
|
||||||
|
|
||||||
% Copyright (C) 2006-2011 Dynare Team
|
% Copyright (C) 2006-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -49,7 +49,7 @@ function [Pstar,Pinf] = compute_Pinf_Pstar(mf,T,R,Q,qz_criterium, restrict_colum
|
||||||
|
|
||||||
np = size(T,1);
|
np = size(T,1);
|
||||||
|
|
||||||
% perform Kitagawa transformation
|
% perform Kitagawa transformation
|
||||||
[QT,ST] = schur(T);
|
[QT,ST] = schur(T);
|
||||||
e1 = abs(ordeig(ST)) > 2-qz_criterium;
|
e1 = abs(ordeig(ST)) > 2-qz_criterium;
|
||||||
[QT,ST] = ordschur(QT,ST,e1);
|
[QT,ST] = ordschur(QT,ST,e1);
|
||||||
|
@ -112,4 +112,3 @@ end
|
||||||
|
|
||||||
Pinf = QT*Pinf*QT';
|
Pinf = QT*Pinf*QT';
|
||||||
Pstar = QT*Pstar*QT';
|
Pstar = QT*Pstar*QT';
|
||||||
|
|
||||||
|
|
|
@ -1,21 +1,21 @@
|
||||||
function [posterior_mean,posterior_covariance,posterior_mode,posterior_kernel_at_the_mode] = compute_mh_covariance_matrix()
|
function [posterior_mean,posterior_covariance,posterior_mode,posterior_kernel_at_the_mode] = compute_mh_covariance_matrix()
|
||||||
% Estimation of the posterior covariance matrix, posterior mean, posterior mode and evaluation of the posterior kernel at the
|
% Estimation of the posterior covariance matrix, posterior mean, posterior mode and evaluation of the posterior kernel at the
|
||||||
% estimated mode, using the draws from a metropolis-hastings. The estimated posterior mode and covariance matrix are saved in
|
% estimated mode, using the draws from a metropolis-hastings. The estimated posterior mode and covariance matrix are saved in
|
||||||
% a file <M_.fname>_mh_mode.mat.
|
% a file <M_.fname>_mh_mode.mat.
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% None.
|
% None.
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% o posterior_mean [double] (n*1) vector, posterior expectation of the parameters.
|
% o posterior_mean [double] (n*1) vector, posterior expectation of the parameters.
|
||||||
% o posterior_covariance [double] (n*n) matrix, posterior covariance of the parameters (computed from previous metropolis hastings).
|
% o posterior_covariance [double] (n*n) matrix, posterior covariance of the parameters (computed from previous metropolis hastings).
|
||||||
% o posterior_mode [double] (n*1) vector, posterior mode of the parameters.
|
% o posterior_mode [double] (n*1) vector, posterior mode of the parameters.
|
||||||
% o posterior_kernel_at_the_mode [double] scalar.
|
% o posterior_kernel_at_the_mode [double] scalar.
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% None.
|
% None.
|
||||||
|
|
||||||
% Copyright (C) 2006-2011 Dynare Team
|
% Copyright (C) 2006-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -62,7 +62,7 @@ offset = 0;
|
||||||
for b=1:nblck
|
for b=1:nblck
|
||||||
first_line = FirstLine;
|
first_line = FirstLine;
|
||||||
for n = FirstMhFile:TotalNumberOfMhFiles
|
for n = FirstMhFile:TotalNumberOfMhFiles
|
||||||
load([ BaseName '_mh' int2str(n) '_blck' int2str(b) '.mat'],'x2','logpo2');
|
load([ BaseName '_mh' int2str(n) '_blck' int2str(b) '.mat'],'x2','logpo2');
|
||||||
[tmp,idx] = max(logpo2);
|
[tmp,idx] = max(logpo2);
|
||||||
if tmp>posterior_kernel_at_the_mode
|
if tmp>posterior_kernel_at_the_mode
|
||||||
posterior_kernel_at_the_mode = tmp;
|
posterior_kernel_at_the_mode = tmp;
|
||||||
|
|
|
@ -4,14 +4,14 @@ function moments=compute_model_moments(dr,M_,options_)
|
||||||
% dr: structure describing model solution
|
% dr: structure describing model solution
|
||||||
% M_: structure of Dynare options
|
% M_: structure of Dynare options
|
||||||
% options_
|
% options_
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% moments: a cell array containing requested moments
|
% moments: a cell array containing requested moments
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2008-2009 Dynare Team
|
% Copyright (C) 2008-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
|
|
@ -2,21 +2,21 @@ function oo_ = compute_moments_varendo(type,options_,M_,oo_,var_list_)
|
||||||
% Computes the second order moments (autocorrelation function, covariance
|
% Computes the second order moments (autocorrelation function, covariance
|
||||||
% matrix and variance decomposition) distributions for all the endogenous variables selected in
|
% matrix and variance decomposition) distributions for all the endogenous variables selected in
|
||||||
% var_list_. The results are saved in oo_
|
% var_list_. The results are saved in oo_
|
||||||
%
|
%
|
||||||
% INPUTS:
|
% INPUTS:
|
||||||
% type [string] 'posterior' or 'prior'
|
% type [string] 'posterior' or 'prior'
|
||||||
% options_ [structure] Dynare structure.
|
% options_ [structure] Dynare structure.
|
||||||
% M_ [structure] Dynare structure (related to model definition).
|
% M_ [structure] Dynare structure (related to model definition).
|
||||||
% oo_ [structure] Dynare structure (results).
|
% oo_ [structure] Dynare structure (results).
|
||||||
% var_list_ [string] Array of string with endogenous variable names.
|
% var_list_ [string] Array of string with endogenous variable names.
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% oo_ [structure] Dynare structure (results).
|
% oo_ [structure] Dynare structure (results).
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2008-2010 Dynare Team
|
% Copyright (C) 2008-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -31,7 +31,7 @@ function oo_ = compute_moments_varendo(type,options_,M_,oo_,var_list_)
|
||||||
% GNU General Public License for more details.
|
% GNU General Public License for more details.
|
||||||
%
|
%
|
||||||
% You should have received a copy of the GNU General Public License
|
% You should have received a copy of the GNU General Public License
|
||||||
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
|
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
|
||||||
|
|
||||||
|
|
||||||
fprintf('Estimation::compute_moments_varendo: I''m computing endogenous moments (this may take a while)... ');
|
fprintf('Estimation::compute_moments_varendo: I''m computing endogenous moments (this may take a while)... ');
|
||||||
|
|
|
@ -1,6 +1,6 @@
|
||||||
function overallacceptanceratio = compute_overall_acceptance_ratio(MetropolisFolder, ModelName)
|
function overallacceptanceratio = compute_overall_acceptance_ratio(MetropolisFolder, ModelName)
|
||||||
|
|
||||||
% Copyright (C) 2013 Dynare Team
|
% Copyright (C) 2013-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -24,12 +24,12 @@ n = length(mh_history_files);
|
||||||
|
|
||||||
load([BaseName '_mh_history_' num2str(0)]);
|
load([BaseName '_mh_history_' num2str(0)]);
|
||||||
TotalNumberOfDraws = record.MhDraws(end,1);
|
TotalNumberOfDraws = record.MhDraws(end,1);
|
||||||
TotalNumberOfAcceptedProposals = record.AcceptanceRatio*record.MhDraws(end,1);
|
TotalNumberOfAcceptedProposals = record.AcceptanceRatio*record.MhDraws(end,1);
|
||||||
|
|
||||||
for i=2:n
|
for i=2:n
|
||||||
load([BaseName '_mh_history_' num2str(i-1)]);
|
load([BaseName '_mh_history_' num2str(i-1)]);
|
||||||
TotalNumberOfDraws = TotalNumberOfDraws + record.MhDraws(end,1);
|
TotalNumberOfDraws = TotalNumberOfDraws + record.MhDraws(end,1);
|
||||||
TotalNumberOfAcceptedProposals = TotalNumberOfAcceptedProposals + record.AcceptanceRatio*record.MhDraws(end,1);
|
TotalNumberOfAcceptedProposals = TotalNumberOfAcceptedProposals + record.AcceptanceRatio*record.MhDraws(end,1);
|
||||||
end
|
end
|
||||||
|
|
||||||
overallacceptanceratio = TotalNumberOfAcceptedProposals/TotalNumberOfDraws;
|
overallacceptanceratio = TotalNumberOfAcceptedProposals/TotalNumberOfDraws;
|
|
@ -18,7 +18,7 @@ function [trend_addition, trend_coeff]=compute_trend_coefficients(M_,DynareOptio
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2014 Dynare Team
|
% Copyright (C) 2014-2016 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
|
|
@ -1,14 +1,14 @@
|
||||||
function ConditionalVarianceDecomposition = conditional_variance_decomposition(StateSpaceModel, Steps, SubsetOfVariables,sigma_e_is_diagonal)
|
function ConditionalVarianceDecomposition = conditional_variance_decomposition(StateSpaceModel, Steps, SubsetOfVariables,sigma_e_is_diagonal)
|
||||||
% This function computes the conditional variance decomposition of a given state space model
|
% This function computes the conditional variance decomposition of a given state space model
|
||||||
% for a subset of endogenous variables.
|
% for a subset of endogenous variables.
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% StateSpaceModel [structure] Specification of the state space model.
|
% StateSpaceModel [structure] Specification of the state space model.
|
||||||
% Steps [integer] 1*h vector of dates.
|
% Steps [integer] 1*h vector of dates.
|
||||||
% SubsetOfVariables [integer] 1*q vector of indices.
|
% SubsetOfVariables [integer] 1*q vector of indices.
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% ConditionalVarianceDecomposition [double] [n h p] array, where
|
% ConditionalVarianceDecomposition [double] [n h p] array, where
|
||||||
% n is equal to length(SubsetOfVariables)
|
% n is equal to length(SubsetOfVariables)
|
||||||
% h is the number of Steps
|
% h is the number of Steps
|
||||||
% p is the number of state innovations and
|
% p is the number of state innovations and
|
||||||
|
@ -16,7 +16,7 @@ function ConditionalVarianceDecomposition = conditional_variance_decomposition(S
|
||||||
%
|
%
|
||||||
% [1] In this version, absence of measurement errors is assumed...
|
% [1] In this version, absence of measurement errors is assumed...
|
||||||
|
|
||||||
% Copyright (C) 2010-2011 Dynare Team
|
% Copyright (C) 2010-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -77,7 +77,7 @@ for h = 1:length(Steps)
|
||||||
SumOfVariances(:,h) = sum(ConditionalVariance(:,h,:),3);
|
SumOfVariances(:,h) = sum(ConditionalVariance(:,h,:),3);
|
||||||
end
|
end
|
||||||
|
|
||||||
ConditionalVarianceDecomposition = zeros(NumberOfVariables,length(Steps),number_of_state_innovations);
|
ConditionalVarianceDecomposition = zeros(NumberOfVariables,length(Steps),number_of_state_innovations);
|
||||||
for i=1:number_of_state_innovations
|
for i=1:number_of_state_innovations
|
||||||
for h = 1:length(Steps)
|
for h = 1:length(Steps)
|
||||||
ConditionalVarianceDecomposition(:,h,i) = squeeze(ConditionalVariance(:,h,i))./SumOfVariances(:,h);
|
ConditionalVarianceDecomposition(:,h,i) = squeeze(ConditionalVariance(:,h,i))./SumOfVariances(:,h);
|
||||||
|
|
|
@ -9,7 +9,7 @@ function oo_ = ...
|
||||||
% dname [string] directory name where to save
|
% dname [string] directory name where to save
|
||||||
% fname [string] name of the mod-file
|
% fname [string] name of the mod-file
|
||||||
% Steps [integers] horizons at which to conduct decomposition
|
% Steps [integers] horizons at which to conduct decomposition
|
||||||
% exonames [string] (n_exo*char_length) character array with names of exogenous variables
|
% exonames [string] (n_exo*char_length) character array with names of exogenous variables
|
||||||
% exo [string] name of current exogenous
|
% exo [string] name of current exogenous
|
||||||
% variable
|
% variable
|
||||||
% var_list [string] (n_endo*char_length) character array with name
|
% var_list [string] (n_endo*char_length) character array with name
|
||||||
|
@ -23,7 +23,7 @@ function oo_ = ...
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% oo_ [structure] Dynare structure where the results are saved.
|
% oo_ [structure] Dynare structure where the results are saved.
|
||||||
|
|
||||||
% Copyright (C) 2009-2015 Dynare Team
|
% Copyright (C) 2009-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -106,7 +106,7 @@ for i=1:length(Steps)
|
||||||
p_density(:,:,i) = pp_density;
|
p_density(:,:,i) = pp_density;
|
||||||
else
|
else
|
||||||
[pp_mean, pp_median, pp_var, hpd_interval, pp_deciles] = ...
|
[pp_mean, pp_median, pp_var, hpd_interval, pp_deciles] = ...
|
||||||
posterior_moments(tmp(:,i),0,mh_conf_sig);
|
posterior_moments(tmp(:,i),0,mh_conf_sig);
|
||||||
end
|
end
|
||||||
p_mean(i) = pp_mean;
|
p_mean(i) = pp_mean;
|
||||||
p_median(i) = pp_median;
|
p_median(i) = pp_median;
|
||||||
|
|
|
@ -1,14 +1,14 @@
|
||||||
function oo_ = McMCDiagnostics(options_, estim_params_, M_, oo_)
|
function oo_ = McMCDiagnostics(options_, estim_params_, M_, oo_)
|
||||||
% function McMCDiagnostics
|
% function McMCDiagnostics
|
||||||
% Computes convergence tests
|
% Computes convergence tests
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% options_ [structure]
|
% options_ [structure]
|
||||||
% estim_params_ [structure]
|
% estim_params_ [structure]
|
||||||
% M_ [structure]
|
% M_ [structure]
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% oo_ [structure]
|
% oo_ [structure]
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
@ -82,12 +82,12 @@ param_name_tex=[];
|
||||||
for jj=1:npar
|
for jj=1:npar
|
||||||
if options_.TeX
|
if options_.TeX
|
||||||
[par_name_temp,par_name_tex_temp]=get_the_name(jj,options_.TeX,M_,estim_params_,options_);
|
[par_name_temp,par_name_tex_temp]=get_the_name(jj,options_.TeX,M_,estim_params_,options_);
|
||||||
param_name = strvcat(param_name,par_name_temp);
|
param_name = strvcat(param_name,par_name_temp);
|
||||||
par_name_tex_temp = strrep(par_name_tex_temp,'$','');
|
par_name_tex_temp = strrep(par_name_tex_temp,'$','');
|
||||||
param_name_tex = strvcat(param_name_tex,par_name_tex_temp);
|
param_name_tex = strvcat(param_name_tex,par_name_tex_temp);
|
||||||
else
|
else
|
||||||
[par_name_temp]=get_the_name(jj,options_.TeX,M_,estim_params_,options_);
|
[par_name_temp]=get_the_name(jj,options_.TeX,M_,estim_params_,options_);
|
||||||
param_name = strvcat(param_name,par_name_temp);
|
param_name = strvcat(param_name,par_name_temp);
|
||||||
end
|
end
|
||||||
Draws = GetAllPosteriorDraws(jj,FirstMhFile,FirstLine,TotalNumberOfMhFiles,NumberOfDraws);
|
Draws = GetAllPosteriorDraws(jj,FirstMhFile,FirstLine,TotalNumberOfMhFiles,NumberOfDraws);
|
||||||
Draws = reshape(Draws,[NumberOfDraws nblck]);
|
Draws = reshape(Draws,[NumberOfDraws nblck]);
|
||||||
|
@ -101,7 +101,7 @@ end
|
||||||
my_title='MCMC Inefficiency factors per block';
|
my_title='MCMC Inefficiency factors per block';
|
||||||
IFAC_header='Parameter';
|
IFAC_header='Parameter';
|
||||||
IFAC_header_tex='Parameter';
|
IFAC_header_tex='Parameter';
|
||||||
for j=1:nblck,
|
for j=1:nblck
|
||||||
IFAC_header = char(IFAC_header,['Block ' int2str(j)]);
|
IFAC_header = char(IFAC_header,['Block ' int2str(j)]);
|
||||||
IFAC_header_tex = char(IFAC_header_tex,['Block~' int2str(j)]);
|
IFAC_header_tex = char(IFAC_header_tex,['Block~' int2str(j)]);
|
||||||
end
|
end
|
||||||
|
@ -139,9 +139,9 @@ if nblck == 1 % Brooks and Gelman tests need more than one block
|
||||||
first_obs_end_sample = first_obs_begin_sample+round(options_.convergence.geweke.geweke_interval(2)*NumberOfDraws*(1-options_.mh_drop));
|
first_obs_end_sample = first_obs_begin_sample+round(options_.convergence.geweke.geweke_interval(2)*NumberOfDraws*(1-options_.mh_drop));
|
||||||
param_name=[];
|
param_name=[];
|
||||||
if options_.TeX
|
if options_.TeX
|
||||||
param_name_tex=[];
|
param_name_tex=[];
|
||||||
end
|
end
|
||||||
for jj=1:npar
|
for jj=1:npar
|
||||||
if options_.TeX
|
if options_.TeX
|
||||||
[param_name_temp, param_name_tex_temp]= get_the_name(jj,options_.TeX,M_,estim_params_,options_);
|
[param_name_temp, param_name_tex_temp]= get_the_name(jj,options_.TeX,M_,estim_params_,options_);
|
||||||
param_name_tex = strvcat(param_name_tex,strrep(param_name_tex_temp,'$',''));
|
param_name_tex = strvcat(param_name_tex,strrep(param_name_tex_temp,'$',''));
|
||||||
|
@ -152,7 +152,7 @@ if nblck == 1 % Brooks and Gelman tests need more than one block
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
fprintf('\nGeweke (1992) Convergence Tests, based on means of draws %d to %d vs %d to %d.\n',first_obs_begin_sample,last_obs_begin_sample,first_obs_end_sample,NumberOfDraws);
|
fprintf('\nGeweke (1992) Convergence Tests, based on means of draws %d to %d vs %d to %d.\n',first_obs_begin_sample,last_obs_begin_sample,first_obs_end_sample,NumberOfDraws);
|
||||||
fprintf('p-values are for Chi2-test for equality of means.\n');
|
fprintf('p-values are for Chi2-test for equality of means.\n');
|
||||||
Geweke_header=char('Parameter', 'Post. Mean', 'Post. Std', 'p-val No Taper');
|
Geweke_header=char('Parameter', 'Post. Mean', 'Post. Std', 'p-val No Taper');
|
||||||
for ii=1:length(options_.convergence.geweke.taper_steps)
|
for ii=1:length(options_.convergence.geweke.taper_steps)
|
||||||
Geweke_header=char(Geweke_header,['p-val ' num2str(options_.convergence.geweke.taper_steps(ii)),'% Taper']);
|
Geweke_header=char(Geweke_header,['p-val ' num2str(options_.convergence.geweke.taper_steps(ii)),'% Taper']);
|
||||||
|
@ -168,12 +168,12 @@ if nblck == 1 % Brooks and Gelman tests need more than one block
|
||||||
end
|
end
|
||||||
[results_vec, results_struct] = geweke_moments(param_draws,options_);
|
[results_vec, results_struct] = geweke_moments(param_draws,options_);
|
||||||
convergence_diagnostics_geweke(jj,:)=results_vec;
|
convergence_diagnostics_geweke(jj,:)=results_vec;
|
||||||
|
|
||||||
param_draws1 = param_draws(first_obs_begin_sample:last_obs_begin_sample,:);
|
param_draws1 = param_draws(first_obs_begin_sample:last_obs_begin_sample,:);
|
||||||
param_draws2 = param_draws(first_obs_end_sample:end,:);
|
param_draws2 = param_draws(first_obs_end_sample:end,:);
|
||||||
[results_vec1] = geweke_moments(param_draws1,options_);
|
[results_vec1] = geweke_moments(param_draws1,options_);
|
||||||
[results_vec2] = geweke_moments(param_draws2,options_);
|
[results_vec2] = geweke_moments(param_draws2,options_);
|
||||||
|
|
||||||
results_struct = geweke_chi2_test(results_vec1,results_vec2,results_struct,options_);
|
results_struct = geweke_chi2_test(results_vec1,results_vec2,results_struct,options_);
|
||||||
eval(['oo_.convergence.geweke.',param_name(jj,:),'=results_struct;'])
|
eval(['oo_.convergence.geweke.',param_name(jj,:),'=results_struct;'])
|
||||||
datamat(jj,:)=[results_struct.posteriormean,results_struct.posteriorstd,results_struct.prob_chi2_test];
|
datamat(jj,:)=[results_struct.posteriormean,results_struct.posteriorstd,results_struct.prob_chi2_test];
|
||||||
|
@ -190,22 +190,22 @@ if nblck == 1 % Brooks and Gelman tests need more than one block
|
||||||
headers = char(Geweke_tex_header);
|
headers = char(Geweke_tex_header);
|
||||||
lh = size(param_name_tex,2)+2;
|
lh = size(param_name_tex,2)+2;
|
||||||
my_title=sprintf('Geweke (1992) Convergence Tests, based on means of draws %d to %d vs %d to %d. p-values are for $\\\\chi^2$-test for equality of means.',first_obs_begin_sample,last_obs_begin_sample,first_obs_end_sample,NumberOfDraws);
|
my_title=sprintf('Geweke (1992) Convergence Tests, based on means of draws %d to %d vs %d to %d. p-values are for $\\\\chi^2$-test for equality of means.',first_obs_begin_sample,last_obs_begin_sample,first_obs_end_sample,NumberOfDraws);
|
||||||
dyn_latex_table(M_,options_,my_title,'geweke',headers,param_name_tex,datamat,lh,12,4,additional_header);
|
dyn_latex_table(M_,options_,my_title,'geweke',headers,param_name_tex,datamat,lh,12,4,additional_header);
|
||||||
end
|
end
|
||||||
skipline(2);
|
skipline(2);
|
||||||
|
|
||||||
if options_.convergence.rafterylewis.indicator
|
if options_.convergence.rafterylewis.indicator
|
||||||
if any(options_.convergence.rafterylewis.qrs<0) || any(options_.convergence.rafterylewis.qrs>1) || length(options_.convergence.rafterylewis.qrs)~=3 ...
|
if any(options_.convergence.rafterylewis.qrs<0) || any(options_.convergence.rafterylewis.qrs>1) || length(options_.convergence.rafterylewis.qrs)~=3 ...
|
||||||
|| (options_.convergence.rafterylewis.qrs(1)-options_.convergence.rafterylewis.qrs(2)<=0)
|
|| (options_.convergence.rafterylewis.qrs(1)-options_.convergence.rafterylewis.qrs(2)<=0)
|
||||||
fprintf('\nCONVERGENCE DIAGNOSTICS: Invalid option for raftery_lewis_qrs. Using the default of [0.025 0.005 0.95].\n')
|
fprintf('\nCONVERGENCE DIAGNOSTICS: Invalid option for raftery_lewis_qrs. Using the default of [0.025 0.005 0.95].\n')
|
||||||
options_.convergence.rafterylewis.qrs=[0.025 0.005 0.95];
|
options_.convergence.rafterylewis.qrs=[0.025 0.005 0.95];
|
||||||
end
|
end
|
||||||
Raftery_Lewis_q=options_.convergence.rafterylewis.qrs(1);
|
Raftery_Lewis_q=options_.convergence.rafterylewis.qrs(1);
|
||||||
Raftery_Lewis_r=options_.convergence.rafterylewis.qrs(2);
|
Raftery_Lewis_r=options_.convergence.rafterylewis.qrs(2);
|
||||||
Raftery_Lewis_s=options_.convergence.rafterylewis.qrs(3);
|
Raftery_Lewis_s=options_.convergence.rafterylewis.qrs(3);
|
||||||
oo_.Raftery_Lewis = raftery_lewis(x2,Raftery_Lewis_q,Raftery_Lewis_r,Raftery_Lewis_s);
|
oo_.Raftery_Lewis = raftery_lewis(x2,Raftery_Lewis_q,Raftery_Lewis_r,Raftery_Lewis_s);
|
||||||
oo_.Raftery_Lewis.parameter_names=param_name;
|
oo_.Raftery_Lewis.parameter_names=param_name;
|
||||||
my_title=sprintf('Raftery/Lewis (1992) Convergence Diagnostics, based on quantile q=%4.3f with precision r=%4.3f with probability s=%4.3f.',Raftery_Lewis_q,Raftery_Lewis_r,Raftery_Lewis_s);
|
my_title=sprintf('Raftery/Lewis (1992) Convergence Diagnostics, based on quantile q=%4.3f with precision r=%4.3f with probability s=%4.3f.',Raftery_Lewis_q,Raftery_Lewis_r,Raftery_Lewis_s);
|
||||||
headers = char('Variables','M (burn-in)','N (req. draws)','N+M (total draws)','k (thinning)');
|
headers = char('Variables','M (burn-in)','N (req. draws)','N+M (total draws)','k (thinning)');
|
||||||
|
|
||||||
raftery_data_mat=[oo_.Raftery_Lewis.M_burn,oo_.Raftery_Lewis.N_prec,oo_.Raftery_Lewis.N_total,oo_.Raftery_Lewis.k_thin];
|
raftery_data_mat=[oo_.Raftery_Lewis.M_burn,oo_.Raftery_Lewis.N_prec,oo_.Raftery_Lewis.N_total,oo_.Raftery_Lewis.k_thin];
|
||||||
|
@ -217,15 +217,15 @@ if nblck == 1 % Brooks and Gelman tests need more than one block
|
||||||
labels_Raftery_Lewis_tex=char(param_name_tex,'Maximum');
|
labels_Raftery_Lewis_tex=char(param_name_tex,'Maximum');
|
||||||
lh = size(labels_Raftery_Lewis_tex,2)+2;
|
lh = size(labels_Raftery_Lewis_tex,2)+2;
|
||||||
dyn_latex_table(M_,options_,my_title,'raftery_lewis',headers,labels_Raftery_Lewis_tex,raftery_data_mat,lh,10,0);
|
dyn_latex_table(M_,options_,my_title,'raftery_lewis',headers,labels_Raftery_Lewis_tex,raftery_data_mat,lh,10,0);
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
return;
|
return
|
||||||
end
|
end
|
||||||
|
|
||||||
Origin = 1000;
|
Origin = 1000;
|
||||||
StepSize = ceil((NumberOfDraws-Origin)/100);% So that the computational time does not
|
StepSize = ceil((NumberOfDraws-Origin)/100);% So that the computational time does not
|
||||||
ALPHA = 0.2; % increase too much with the number of simulations.
|
ALPHA = 0.2; % increase too much with the number of simulations.
|
||||||
time = 1:NumberOfDraws;
|
time = 1:NumberOfDraws;
|
||||||
xx = Origin:StepSize:NumberOfDraws;
|
xx = Origin:StepSize:NumberOfDraws;
|
||||||
NumberOfLines = length(xx);
|
NumberOfLines = length(xx);
|
||||||
|
@ -262,7 +262,7 @@ localVars.M_ = M_;
|
||||||
|
|
||||||
|
|
||||||
% Like sequential execution!
|
% Like sequential execution!
|
||||||
if isnumeric(options_.parallel),
|
if isnumeric(options_.parallel)
|
||||||
fout = McMCDiagnostics_core(localVars,1,npar,0);
|
fout = McMCDiagnostics_core(localVars,1,npar,0);
|
||||||
UDIAG = fout.UDIAG;
|
UDIAG = fout.UDIAG;
|
||||||
clear fout
|
clear fout
|
||||||
|
@ -273,19 +273,19 @@ else
|
||||||
ModelName = [ModelName '_bvar'];
|
ModelName = [ModelName '_bvar'];
|
||||||
end
|
end
|
||||||
NamFileInput={[M_.dname '/metropolis/'],[ModelName '_mh*_blck*.mat']};
|
NamFileInput={[M_.dname '/metropolis/'],[ModelName '_mh*_blck*.mat']};
|
||||||
|
|
||||||
[fout, nBlockPerCPU, totCPU] = masterParallel(options_.parallel, 1, npar,NamFileInput,'McMCDiagnostics_core', localVars, [], options_.parallel_info);
|
[fout, nBlockPerCPU, totCPU] = masterParallel(options_.parallel, 1, npar,NamFileInput,'McMCDiagnostics_core', localVars, [], options_.parallel_info);
|
||||||
UDIAG = fout(1).UDIAG;
|
UDIAG = fout(1).UDIAG;
|
||||||
for j=2:totCPU,
|
for j=2:totCPU
|
||||||
UDIAG = cat(3,UDIAG ,fout(j).UDIAG);
|
UDIAG = cat(3,UDIAG ,fout(j).UDIAG);
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
|
|
||||||
UDIAG(:,[2 4 6],:) = UDIAG(:,[2 4 6],:)/nblck;
|
UDIAG(:,[2 4 6],:) = UDIAG(:,[2 4 6],:)/nblck;
|
||||||
skipline()
|
skipline()
|
||||||
clear pmet temp moyenne CSUP CINF csup cinf n linea iter tmp;
|
clear pmet temp moyenne CSUP CINF csup cinf n linea iter tmp;
|
||||||
pages = floor(npar/3);
|
pages = floor(npar/3);
|
||||||
k = 0;
|
k = 0;
|
||||||
for i = 1:pages
|
for i = 1:pages
|
||||||
h=dyn_figure(options_.nodisplay,'Name','MCMC univariate convergence diagnostic (Brooks and Gelman,1998)');
|
h=dyn_figure(options_.nodisplay,'Name','MCMC univariate convergence diagnostic (Brooks and Gelman,1998)');
|
||||||
boxplot = 1;
|
boxplot = 1;
|
||||||
|
@ -296,12 +296,12 @@ for i = 1:pages
|
||||||
if crit == 1
|
if crit == 1
|
||||||
plt1 = UDIAG(:,1,k);
|
plt1 = UDIAG(:,1,k);
|
||||||
plt2 = UDIAG(:,2,k);
|
plt2 = UDIAG(:,2,k);
|
||||||
namnam = [nam , ' (Interval)'];
|
namnam = [nam , ' (Interval)'];
|
||||||
elseif crit == 2
|
elseif crit == 2
|
||||||
plt1 = UDIAG(:,3,k);
|
plt1 = UDIAG(:,3,k);
|
||||||
plt2 = UDIAG(:,4,k);
|
plt2 = UDIAG(:,4,k);
|
||||||
namnam = [nam , ' (m2)'];
|
namnam = [nam , ' (m2)'];
|
||||||
elseif crit == 3
|
elseif crit == 3
|
||||||
plt1 = UDIAG(:,5,k);
|
plt1 = UDIAG(:,5,k);
|
||||||
plt2 = UDIAG(:,6,k);
|
plt2 = UDIAG(:,6,k);
|
||||||
namnam = [nam , ' (m3)'];
|
namnam = [nam , ' (m3)'];
|
||||||
|
@ -330,7 +330,7 @@ for i = 1:pages
|
||||||
fprintf(fidTeX,'\\begin{figure}[H]\n');
|
fprintf(fidTeX,'\\begin{figure}[H]\n');
|
||||||
for jj = 1:size(NAMES,1)
|
for jj = 1:size(NAMES,1)
|
||||||
fprintf(fidTeX,'\\psfrag{%s}[1][][0.5][0]{%s}\n',deblank(NAMES(jj,:)),deblank(TEXNAMES(jj,:)));
|
fprintf(fidTeX,'\\psfrag{%s}[1][][0.5][0]{%s}\n',deblank(NAMES(jj,:)),deblank(TEXNAMES(jj,:)));
|
||||||
end
|
end
|
||||||
fprintf(fidTeX,'\\centering \n');
|
fprintf(fidTeX,'\\centering \n');
|
||||||
fprintf(fidTeX,'\\includegraphics[width=%2.2f\\textwidth]{%s_udiag%s}\n',options_.figures.textwidth*min((boxplot-1)/3,1),[OutputFolder '/' ModelName],int2str(i));
|
fprintf(fidTeX,'\\includegraphics[width=%2.2f\\textwidth]{%s_udiag%s}\n',options_.figures.textwidth*min((boxplot-1)/3,1),[OutputFolder '/' ModelName],int2str(i));
|
||||||
fprintf(fidTeX,'\\caption{Univariate convergence diagnostics for the Metropolis-Hastings.\n');
|
fprintf(fidTeX,'\\caption{Univariate convergence diagnostics for the Metropolis-Hastings.\n');
|
||||||
|
@ -346,7 +346,7 @@ if reste
|
||||||
if reste == 1
|
if reste == 1
|
||||||
nr = 3;
|
nr = 3;
|
||||||
nc = 1;
|
nc = 1;
|
||||||
elseif reste == 2;
|
elseif reste == 2
|
||||||
nr = 2;
|
nr = 2;
|
||||||
nc = 3;
|
nc = 3;
|
||||||
end
|
end
|
||||||
|
@ -359,12 +359,12 @@ if reste
|
||||||
if crit == 1
|
if crit == 1
|
||||||
plt1 = UDIAG(:,1,k);
|
plt1 = UDIAG(:,1,k);
|
||||||
plt2 = UDIAG(:,2,k);
|
plt2 = UDIAG(:,2,k);
|
||||||
namnam = [nam , ' (Interval)'];
|
namnam = [nam , ' (Interval)'];
|
||||||
elseif crit == 2
|
elseif crit == 2
|
||||||
plt1 = UDIAG(:,3,k);
|
plt1 = UDIAG(:,3,k);
|
||||||
plt2 = UDIAG(:,4,k);
|
plt2 = UDIAG(:,4,k);
|
||||||
namnam = [nam , ' (m2)'];
|
namnam = [nam , ' (m2)'];
|
||||||
elseif crit == 3
|
elseif crit == 3
|
||||||
plt1 = UDIAG(:,5,k);
|
plt1 = UDIAG(:,5,k);
|
||||||
plt2 = UDIAG(:,6,k);
|
plt2 = UDIAG(:,6,k);
|
||||||
namnam = [nam , ' (m3)'];
|
namnam = [nam , ' (m3)'];
|
||||||
|
@ -391,9 +391,9 @@ if reste
|
||||||
dyn_saveas(h,[ OutputFolder '/' ModelName '_udiag' int2str(pages+1)],options_.nodisplay,options_.graph_format);
|
dyn_saveas(h,[ OutputFolder '/' ModelName '_udiag' int2str(pages+1)],options_.nodisplay,options_.graph_format);
|
||||||
if TeX && any(strcmp('eps',cellstr(options_.graph_format)))
|
if TeX && any(strcmp('eps',cellstr(options_.graph_format)))
|
||||||
fprintf(fidTeX,'\\begin{figure}[H]\n');
|
fprintf(fidTeX,'\\begin{figure}[H]\n');
|
||||||
for jj = 1:size(NAMES,1);
|
for jj = 1:size(NAMES,1)
|
||||||
fprintf(fidTeX,'\\psfrag{%s}[1][][0.5][0]{%s}\n',deblank(NAMES(jj,:)),deblank(TEXNAMES(jj,:)));
|
fprintf(fidTeX,'\\psfrag{%s}[1][][0.5][0]{%s}\n',deblank(NAMES(jj,:)),deblank(TEXNAMES(jj,:)));
|
||||||
end
|
end
|
||||||
fprintf(fidTeX,'\\centering \n');
|
fprintf(fidTeX,'\\centering \n');
|
||||||
fprintf(fidTeX,'\\includegraphics[width=%2.2f\\textwidth]{%s_udiag%s}\n',options_.figures.textwidth*min((boxplot-1)/nc,1),[OutputFolder '/' ModelName],int2str(pages+1));
|
fprintf(fidTeX,'\\includegraphics[width=%2.2f\\textwidth]{%s_udiag%s}\n',options_.figures.textwidth*min((boxplot-1)/nc,1),[OutputFolder '/' ModelName],int2str(pages+1));
|
||||||
if reste == 2
|
if reste == 2
|
||||||
|
@ -435,7 +435,7 @@ for b = 1:nblck
|
||||||
end
|
end
|
||||||
clear logpo2;
|
clear logpo2;
|
||||||
tmp(:,2) = kron(transpose(1:nblck),ones(NumberOfDraws,1));
|
tmp(:,2) = kron(transpose(1:nblck),ones(NumberOfDraws,1));
|
||||||
tmp(:,3) = kron(ones(nblck,1),time');
|
tmp(:,3) = kron(ones(nblck,1),time');
|
||||||
tmp = sortrows(tmp,1);
|
tmp = sortrows(tmp,1);
|
||||||
ligne = 0;
|
ligne = 0;
|
||||||
for iter = Origin:StepSize:NumberOfDraws
|
for iter = Origin:StepSize:NumberOfDraws
|
||||||
|
@ -454,12 +454,12 @@ for iter = Origin:StepSize:NumberOfDraws
|
||||||
for i=1:nblck
|
for i=1:nblck
|
||||||
pmet = temp(find(temp(:,2)==i));
|
pmet = temp(find(temp(:,2)==i));
|
||||||
MDIAG(ligne,2) = MDIAG(ligne,2) + pmet(csup,1)-pmet(cinf,1);
|
MDIAG(ligne,2) = MDIAG(ligne,2) + pmet(csup,1)-pmet(cinf,1);
|
||||||
moyenne = mean(pmet,1); %% Within mean.
|
moyenne = mean(pmet,1); %% Within mean.
|
||||||
MDIAG(ligne,4) = MDIAG(ligne,4) + sum((pmet(:,1)-moyenne).^2)/(n-1);
|
MDIAG(ligne,4) = MDIAG(ligne,4) + sum((pmet(:,1)-moyenne).^2)/(n-1);
|
||||||
MDIAG(ligne,6) = MDIAG(ligne,6) + sum(abs(pmet(:,1)-moyenne).^3)/(n-1);
|
MDIAG(ligne,6) = MDIAG(ligne,6) + sum(abs(pmet(:,1)-moyenne).^3)/(n-1);
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
MDIAG(:,[2 4 6],:) = MDIAG(:,[2 4 6],:)/nblck;
|
MDIAG(:,[2 4 6],:) = MDIAG(:,[2 4 6],:)/nblck;
|
||||||
|
|
||||||
h = dyn_figure(options_.nodisplay,'Name','Multivariate convergence diagnostic');
|
h = dyn_figure(options_.nodisplay,'Name','Multivariate convergence diagnostic');
|
||||||
boxplot = 1;
|
boxplot = 1;
|
||||||
|
@ -467,12 +467,12 @@ for crit = 1:3
|
||||||
if crit == 1
|
if crit == 1
|
||||||
plt1 = MDIAG(:,1);
|
plt1 = MDIAG(:,1);
|
||||||
plt2 = MDIAG(:,2);
|
plt2 = MDIAG(:,2);
|
||||||
namnam = 'Interval';
|
namnam = 'Interval';
|
||||||
elseif crit == 2
|
elseif crit == 2
|
||||||
plt1 = MDIAG(:,3);
|
plt1 = MDIAG(:,3);
|
||||||
plt2 = MDIAG(:,4);
|
plt2 = MDIAG(:,4);
|
||||||
namnam = 'm2';
|
namnam = 'm2';
|
||||||
elseif crit == 3
|
elseif crit == 3
|
||||||
plt1 = MDIAG(:,5);
|
plt1 = MDIAG(:,5);
|
||||||
plt2 = MDIAG(:,6);
|
plt2 = MDIAG(:,6);
|
||||||
namnam = 'm3';
|
namnam = 'm3';
|
||||||
|
@ -499,7 +499,7 @@ if TeX && any(strcmp('eps',cellstr(options_.graph_format)))
|
||||||
fprintf(fidTeX,'\\begin{figure}[H]\n');
|
fprintf(fidTeX,'\\begin{figure}[H]\n');
|
||||||
for jj = 1:3
|
for jj = 1:3
|
||||||
fprintf(fidTeX,'\\psfrag{%s}[1][][0.5][0]{%s}\n',deblank(NAMES(jj,:)),' ');
|
fprintf(fidTeX,'\\psfrag{%s}[1][][0.5][0]{%s}\n',deblank(NAMES(jj,:)),' ');
|
||||||
end
|
end
|
||||||
fprintf(fidTeX,'\\centering \n');
|
fprintf(fidTeX,'\\centering \n');
|
||||||
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_mdiag}\n',[OutputFolder '/' ModelName]);
|
fprintf(fidTeX,'\\includegraphics[width=0.8\\textwidth]{%s_mdiag}\n',[OutputFolder '/' ModelName]);
|
||||||
fprintf(fidTeX,'\\caption{Multivariate convergence diagnostics for the Metropolis-Hastings.\n');
|
fprintf(fidTeX,'\\caption{Multivariate convergence diagnostics for the Metropolis-Hastings.\n');
|
||||||
|
@ -512,4 +512,3 @@ if TeX && any(strcmp('eps',cellstr(options_.graph_format)))
|
||||||
fprintf(fidTeX,'%% End Of TeX file.');
|
fprintf(fidTeX,'%% End Of TeX file.');
|
||||||
fclose(fidTeX);
|
fclose(fidTeX);
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
|
@ -2,12 +2,12 @@ function myoutput = McMCDiagnostics_core(myinputs,fpar,npar,whoiam, ThisMatlab)
|
||||||
% function myoutput = McMCDiagnostics_core(myinputs,fpar,npar,whoiam, ThisMatlab)
|
% function myoutput = McMCDiagnostics_core(myinputs,fpar,npar,whoiam, ThisMatlab)
|
||||||
% Computes the Brooks/Gelman (1998) convergence diagnostics, both the
|
% Computes the Brooks/Gelman (1998) convergence diagnostics, both the
|
||||||
% parameteric and the non-parameteric versions
|
% parameteric and the non-parameteric versions
|
||||||
%
|
%
|
||||||
% PARALLEL CONTEXT
|
% PARALLEL CONTEXT
|
||||||
% Core functionality for MCMC Diagnostics, which can be parallelized.
|
% Core functionality for MCMC Diagnostics, which can be parallelized.
|
||||||
% See also the comment in posterior_sampler_core.m funtion.
|
% See also the comment in posterior_sampler_core.m funtion.
|
||||||
%
|
%
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% See See the comment in posterior_sampler_core.m funtion.
|
% See See the comment in posterior_sampler_core.m funtion.
|
||||||
|
|
||||||
|
@ -25,15 +25,15 @@ function myoutput = McMCDiagnostics_core(myinputs,fpar,npar,whoiam, ThisMatlab)
|
||||||
%
|
%
|
||||||
% ALGORITHM
|
% ALGORITHM
|
||||||
% Computes part of the convergence diagnostics, the rest is computed in McMCDiagnostics.m .
|
% Computes part of the convergence diagnostics, the rest is computed in McMCDiagnostics.m .
|
||||||
% The methodology and terminology is based on: Brooks/Gelman (1998): General
|
% The methodology and terminology is based on: Brooks/Gelman (1998): General
|
||||||
% Methods for Monitoring Convergence of Iterative Simulations, Journal of Computational
|
% Methods for Monitoring Convergence of Iterative Simulations, Journal of Computational
|
||||||
% and Graphical Statistics, Volume 7, Number 4, Pages 434-455
|
% and Graphical Statistics, Volume 7, Number 4, Pages 434-455
|
||||||
%
|
%
|
||||||
%
|
%
|
||||||
% SPECIAL REQUIREMENTS.
|
% SPECIAL REQUIREMENTS.
|
||||||
% None.
|
% None.
|
||||||
|
|
||||||
% Copyright (C) 2006-2016 Dynare Team
|
% Copyright (C) 2006-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -50,7 +50,7 @@ function myoutput = McMCDiagnostics_core(myinputs,fpar,npar,whoiam, ThisMatlab)
|
||||||
% You should have received a copy of the GNU General Public License
|
% You should have received a copy of the GNU General Public License
|
||||||
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
|
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
|
||||||
|
|
||||||
if nargin<4,
|
if nargin<4
|
||||||
whoiam=0;
|
whoiam=0;
|
||||||
end
|
end
|
||||||
|
|
||||||
|
@ -71,7 +71,7 @@ M_=myinputs.M_;
|
||||||
if whoiam
|
if whoiam
|
||||||
Parallel=myinputs.Parallel;
|
Parallel=myinputs.Parallel;
|
||||||
end
|
end
|
||||||
if ~exist('MetropolisFolder'),
|
if ~exist('MetropolisFolder')
|
||||||
MetropolisFolder = CheckPath('metropolis',M_.dname);
|
MetropolisFolder = CheckPath('metropolis',M_.dname);
|
||||||
end
|
end
|
||||||
|
|
||||||
|
@ -81,16 +81,16 @@ UDIAG = zeros(NumberOfLines,6,npar-fpar+1);
|
||||||
|
|
||||||
if whoiam
|
if whoiam
|
||||||
waitbarString = ['Please wait... McMCDiagnostics (' int2str(fpar) 'of' int2str(npar) ')...'];
|
waitbarString = ['Please wait... McMCDiagnostics (' int2str(fpar) 'of' int2str(npar) ')...'];
|
||||||
if Parallel(ThisMatlab).Local,
|
if Parallel(ThisMatlab).Local
|
||||||
waitbarTitle=['Local '];
|
waitbarTitle=['Local '];
|
||||||
else
|
else
|
||||||
waitbarTitle=[Parallel(ThisMatlab).ComputerName];
|
waitbarTitle=[Parallel(ThisMatlab).ComputerName];
|
||||||
end
|
end
|
||||||
fMessageStatus(0,whoiam,waitbarString, waitbarTitle, Parallel(ThisMatlab));
|
fMessageStatus(0,whoiam,waitbarString, waitbarTitle, Parallel(ThisMatlab));
|
||||||
end
|
end
|
||||||
for j=fpar:npar,
|
for j=fpar:npar
|
||||||
if isoctave
|
if isoctave
|
||||||
if (whoiam==0),
|
if (whoiam==0)
|
||||||
printf(' Parameter %d... ',j);
|
printf(' Parameter %d... ',j);
|
||||||
end
|
end
|
||||||
else
|
else
|
||||||
|
@ -131,13 +131,13 @@ for j=fpar:npar,
|
||||||
end
|
end
|
||||||
end
|
end
|
||||||
if isoctave
|
if isoctave
|
||||||
if (whoiam==0),
|
if (whoiam==0)
|
||||||
printf('Done! \n');
|
printf('Done! \n');
|
||||||
end
|
end
|
||||||
else
|
else
|
||||||
fprintf('Done! \n');
|
fprintf('Done! \n');
|
||||||
end
|
end
|
||||||
if whoiam,
|
if whoiam
|
||||||
waitbarString = [ 'Parameter ' int2str(j) '/' int2str(npar) ' done.'];
|
waitbarString = [ 'Parameter ' int2str(j) '/' int2str(npar) ' done.'];
|
||||||
fMessageStatus((j-fpar+1)/(npar-fpar+1),whoiam,waitbarString, waitbarTitle, Parallel(ThisMatlab))
|
fMessageStatus((j-fpar+1)/(npar-fpar+1),whoiam,waitbarString, waitbarTitle, Parallel(ThisMatlab))
|
||||||
end
|
end
|
||||||
|
|
|
@ -2,7 +2,7 @@ function results_struct = geweke_chi2_test(results1,results2,results_struct,opti
|
||||||
% results_struct = geweke_chi2_test(results1,results2,results_struct,options)
|
% results_struct = geweke_chi2_test(results1,results2,results_struct,options)
|
||||||
% PURPOSE: computes Geweke's chi-squared test for two sets of MCMC sample draws
|
% PURPOSE: computes Geweke's chi-squared test for two sets of MCMC sample draws
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% results1 [1 by (4+n_taper*2) vector] vector with post. mean,
|
% results1 [1 by (4+n_taper*2) vector] vector with post. mean,
|
||||||
% std, NSE_iid, RNE_iid, and tapered NSE and RNE
|
% std, NSE_iid, RNE_iid, and tapered NSE and RNE
|
||||||
% for chain part 1
|
% for chain part 1
|
||||||
|
@ -12,7 +12,7 @@ function results_struct = geweke_chi2_test(results1,results2,results_struct,opti
|
||||||
% results_struct [structure] results structure generated by geweke_moments
|
% results_struct [structure] results structure generated by geweke_moments
|
||||||
% Dynareoptions [structure]
|
% Dynareoptions [structure]
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% results_struct [structure] containing the following fields:
|
% results_struct [structure] containing the following fields:
|
||||||
% pooled_mean Pooled mean of the chain parts, weighted
|
% pooled_mean Pooled mean of the chain parts, weighted
|
||||||
% with precision
|
% with precision
|
||||||
|
@ -26,7 +26,7 @@ function results_struct = geweke_chi2_test(results1,results2,results_struct,opti
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% None.
|
% None.
|
||||||
|
|
||||||
% Copyright (C) 2013 Dynare Team
|
% Copyright (C) 2013-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -49,26 +49,24 @@ function results_struct = geweke_chi2_test(results1,results2,results_struct,opti
|
||||||
% J.M. Bernardo, A.P. Dawid, and A.F.M. Smith (eds.) Proceedings of
|
% J.M. Bernardo, A.P. Dawid, and A.F.M. Smith (eds.) Proceedings of
|
||||||
% the Fourth Valencia International Meeting on Bayesian Statistics,
|
% the Fourth Valencia International Meeting on Bayesian Statistics,
|
||||||
% pp. 169-194, Oxford University Press
|
% pp. 169-194, Oxford University Press
|
||||||
% Geweke (1999): `Using simulation methods for Bayesian econometric models:
|
% Geweke (1999): `Using simulation methods for Bayesian econometric models:
|
||||||
% Inference, development and communication', Econometric Reviews, 18(1),
|
% Inference, development and communication', Econometric Reviews, 18(1),
|
||||||
% 1-73
|
% 1-73
|
||||||
|
|
||||||
% written by: Johannes Pfeifer,
|
% written by: Johannes Pfeifer,
|
||||||
% based on code by James P. LeSage, who in turn
|
% based on code by James P. LeSage, who in turn
|
||||||
% drew on MATLAB programs written by Siddartha Chib
|
% drew on MATLAB programs written by Siddartha Chib
|
||||||
|
|
||||||
for k=1:length(options.convergence.geweke.taper_steps)+1;
|
for k=1:length(options.convergence.geweke.taper_steps)+1
|
||||||
NSE=[results1(:,3+(k-1)*2) results2(:,3+(k-1)*2)];
|
NSE=[results1(:,3+(k-1)*2) results2(:,3+(k-1)*2)];
|
||||||
means=[results1(:,1) results2(:,1)];
|
means=[results1(:,1) results2(:,1)];
|
||||||
diff_Means=means(:,1)-means(:,2);
|
diff_Means=means(:,1)-means(:,2);
|
||||||
sum_of_weights=sum(1./(NSE.^2),2);
|
sum_of_weights=sum(1./(NSE.^2),2);
|
||||||
pooled_mean=sum(means./(NSE.^2),2)./sum_of_weights;
|
pooled_mean=sum(means./(NSE.^2),2)./sum_of_weights;
|
||||||
pooled_NSE=1./sqrt(sum_of_weights);
|
pooled_NSE=1./sqrt(sum_of_weights);
|
||||||
|
test_stat=diff_Means.^2./sum(NSE.^2,2);
|
||||||
test_stat=diff_Means.^2./sum(NSE.^2,2);
|
|
||||||
p = 1-chi2cdf(test_stat,1);
|
p = 1-chi2cdf(test_stat,1);
|
||||||
results_struct.pooled_mean(:,k) = pooled_mean;
|
results_struct.pooled_mean(:,k) = pooled_mean;
|
||||||
results_struct.pooled_nse(:,k) = pooled_NSE;
|
results_struct.pooled_nse(:,k) = pooled_NSE;
|
||||||
results_struct.prob_chi2_test(:,k) = p;
|
results_struct.prob_chi2_test(:,k) = p;
|
||||||
end;
|
end
|
||||||
|
|
||||||
|
|
|
@ -1,13 +1,13 @@
|
||||||
function [results_vec, results_struct] = geweke_moments(draws,Dynareoptions)
|
function [results_vec, results_struct] = geweke_moments(draws,Dynareoptions)
|
||||||
%[results_vec, results_struct] = geweke_moments(draws,Dynareoptions)
|
%[results_vec, results_struct] = geweke_moments(draws,Dynareoptions)
|
||||||
% PURPOSE: computes Gewke's convergence diagnostics NSE and RNE
|
% PURPOSE: computes Gewke's convergence diagnostics NSE and RNE
|
||||||
% (numerical std error and relative numerical efficiencies)
|
% (numerical std error and relative numerical efficiencies)
|
||||||
|
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% draws [ndraws by 1 vector]
|
% draws [ndraws by 1 vector]
|
||||||
% Dynareoptions [structure]
|
% Dynareoptions [structure]
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% results_vec
|
% results_vec
|
||||||
% results_struct [structure] containing the following fields:
|
% results_struct [structure] containing the following fields:
|
||||||
% posteriormean= posterior parameter mean
|
% posteriormean= posterior parameter mean
|
||||||
|
@ -22,7 +22,7 @@ function [results_vec, results_struct] = geweke_moments(draws,Dynareoptions)
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% None.
|
% None.
|
||||||
|
|
||||||
% Copyright (C) 2013 Dynare Team
|
% Copyright (C) 2013-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -44,14 +44,14 @@ function [results_vec, results_struct] = geweke_moments(draws,Dynareoptions)
|
||||||
% J.M. Bernardo, A.P. Dawid, and A.F.M. Smith (eds.) Proceedings of
|
% J.M. Bernardo, A.P. Dawid, and A.F.M. Smith (eds.) Proceedings of
|
||||||
% the Fourth Valencia International Meeting on Bayesian Statistics,
|
% the Fourth Valencia International Meeting on Bayesian Statistics,
|
||||||
% pp. 169-194, Oxford University Press
|
% pp. 169-194, Oxford University Press
|
||||||
% Geweke (1999): `Using simulation methods for Bayesian econometric models:
|
% Geweke (1999): `Using simulation methods for Bayesian econometric models:
|
||||||
% Inference, development and communication', Econometric Reviews, 18(1),
|
% Inference, development and communication', Econometric Reviews, 18(1),
|
||||||
% 1-73
|
% 1-73
|
||||||
% -----------------------------------------------------------------
|
% -----------------------------------------------------------------
|
||||||
|
|
||||||
% written by: Johannes Pfeifer,
|
% written by: Johannes Pfeifer,
|
||||||
% based on code by James P. LeSage, who in turn
|
% based on code by James P. LeSage, who in turn
|
||||||
% drew on MATLAB programs written by Siddartha Chib
|
% drew on MATLAB programs written by Siddartha Chib
|
||||||
|
|
||||||
|
|
||||||
ndraw = size(draws,1);
|
ndraw = size(draws,1);
|
||||||
|
@ -64,10 +64,10 @@ n_draws_used = ns*n_groups; %effective number of draws used after rounding down
|
||||||
|
|
||||||
window_means= zeros(n_groups,1);
|
window_means= zeros(n_groups,1);
|
||||||
window_uncentered_variances= zeros(n_groups,1);
|
window_uncentered_variances= zeros(n_groups,1);
|
||||||
for ig=1:n_groups;
|
for ig=1:n_groups
|
||||||
window_means(ig,1)=sum(draws((ig-1)*ns+1:ig*ns,1))/ns;
|
window_means(ig,1)=sum(draws((ig-1)*ns+1:ig*ns,1))/ns;
|
||||||
window_uncentered_variances(ig,1)=sum(draws((ig-1)*ns+1:ig*ns,1).^2)/ns;
|
window_uncentered_variances(ig,1)=sum(draws((ig-1)*ns+1:ig*ns,1).^2)/ns;
|
||||||
end; %for ig
|
end %for ig
|
||||||
total_mean=mean(window_means);
|
total_mean=mean(window_means);
|
||||||
total_variance=mean(window_uncentered_variances)-total_mean^2;
|
total_variance=mean(window_uncentered_variances)-total_mean^2;
|
||||||
|
|
||||||
|
@ -88,9 +88,9 @@ results_struct.rne_iid = results_vec(1,4);
|
||||||
%get autocovariance of grouped means
|
%get autocovariance of grouped means
|
||||||
centered_window_means=window_means-total_mean;
|
centered_window_means=window_means-total_mean;
|
||||||
autocov_grouped_means=zeros(n_groups,1);
|
autocov_grouped_means=zeros(n_groups,1);
|
||||||
for lag=0:n_groups-1;
|
for lag=0:n_groups-1
|
||||||
autocov_grouped_means(lag+1)=centered_window_means(lag+1:n_groups,1)'*centered_window_means(1:n_groups-lag,1)/100;
|
autocov_grouped_means(lag+1)=centered_window_means(lag+1:n_groups,1)'*centered_window_means(1:n_groups-lag,1)/100;
|
||||||
end;
|
end
|
||||||
|
|
||||||
% numerical standard error with tapered autocovariance functions
|
% numerical standard error with tapered autocovariance functions
|
||||||
for taper_index=1:length(taper_steps)
|
for taper_index=1:length(taper_steps)
|
||||||
|
@ -105,5 +105,4 @@ for taper_index=1:length(taper_steps)
|
||||||
|
|
||||||
eval(['results_struct.nse_taper_',num2str(taper),'= NSE_taper;']);
|
eval(['results_struct.nse_taper_',num2str(taper),'= NSE_taper;']);
|
||||||
eval(['results_struct.rne_taper_',num2str(taper),'= total_variance/(n_draws_used*NSE_taper^2);']);
|
eval(['results_struct.rne_taper_',num2str(taper),'= total_variance/(n_draws_used*NSE_taper^2);']);
|
||||||
end; % end of for mm loop
|
end % end of for mm loop
|
||||||
|
|
||||||
|
|
|
@ -14,12 +14,12 @@ function Ifac = mcmc_ifac(X, Nc)
|
||||||
% ALGORITHM:
|
% ALGORITHM:
|
||||||
% Inefficiency factors are computed as
|
% Inefficiency factors are computed as
|
||||||
% \[
|
% \[
|
||||||
% Ifac = 1 + 2\sum\limits_{i=1}^{Nc} {\hat \rho(i)}
|
% Ifac = 1 + 2\sum\limits_{i=1}^{Nc} {\hat \rho(i)}
|
||||||
% \]
|
% \]
|
||||||
% where $\hat \rho(i)$ denotes the autocorrelation at lag i and the terms
|
% where $\hat \rho(i)$ denotes the autocorrelation at lag i and the terms
|
||||||
% of the sum are truncated using a Parzen window.
|
% of the sum are truncated using a Parzen window.
|
||||||
%
|
%
|
||||||
% For inefficiency factors, see Section 6.1 of Paolo Giordani, Michael Pitt, and Robert Kohn (2011):
|
% For inefficiency factors, see Section 6.1 of Paolo Giordani, Michael Pitt, and Robert Kohn (2011):
|
||||||
% "Bayesian Inference for Time Series State Space Models" in : John Geweke, Gary Koop,
|
% "Bayesian Inference for Time Series State Space Models" in : John Geweke, Gary Koop,
|
||||||
% Herman van Dijk (editors): "The Oxford Handbook of Bayesian
|
% Herman van Dijk (editors): "The Oxford Handbook of Bayesian
|
||||||
% Econometrics", Oxford University Press
|
% Econometrics", Oxford University Press
|
||||||
|
@ -36,7 +36,7 @@ function Ifac = mcmc_ifac(X, Nc)
|
||||||
% consistent covariance matrix estimation", Econometrica, 59(3), p. 817-858
|
% consistent covariance matrix estimation", Econometrica, 59(3), p. 817-858
|
||||||
|
|
||||||
|
|
||||||
% Copyright (C) 2015-16 Dynare Team
|
% Copyright (C) 2015-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -54,7 +54,7 @@ function Ifac = mcmc_ifac(X, Nc)
|
||||||
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
|
% along with Dynare. If not, see <http://www.gnu.org/licenses/>.
|
||||||
|
|
||||||
Nc = floor(min(Nc, length(X)/2));
|
Nc = floor(min(Nc, length(X)/2));
|
||||||
if mod(Nc,2),
|
if mod(Nc,2)
|
||||||
Nc=Nc-1;
|
Nc=Nc-1;
|
||||||
end
|
end
|
||||||
AcorrXSIM = dyn_autocorr(X(:), Nc);
|
AcorrXSIM = dyn_autocorr(X(:), Nc);
|
||||||
|
|
|
@ -1,15 +1,15 @@
|
||||||
function [raftery_lewis] = raftery_lewis(runs,q,r,s)
|
function [raftery_lewis] = raftery_lewis(runs,q,r,s)
|
||||||
% function raftery_lewis = raftery_lewis(runs,q,r,s)
|
% function raftery_lewis = raftery_lewis(runs,q,r,s)
|
||||||
% Computes the convergence diagnostics of Raftery and Lewis (1992), i.e. the
|
% Computes the convergence diagnostics of Raftery and Lewis (1992), i.e. the
|
||||||
% number of draws needed in MCMC to estimate the posterior cdf of the q-quantile
|
% number of draws needed in MCMC to estimate the posterior cdf of the q-quantile
|
||||||
% within an accuracy r with probability s
|
% within an accuracy r with probability s
|
||||||
%
|
%
|
||||||
% Inputs:
|
% Inputs:
|
||||||
% - draws [n_draws by n_var] double matrix of draws from the sampler
|
% - draws [n_draws by n_var] double matrix of draws from the sampler
|
||||||
% - q [scalar] quantile of the quantity of interest
|
% - q [scalar] quantile of the quantity of interest
|
||||||
% - r [scalar] level of desired precision
|
% - r [scalar] level of desired precision
|
||||||
% - s [scalar] probability associated with r
|
% - s [scalar] probability associated with r
|
||||||
%
|
%
|
||||||
% Output:
|
% Output:
|
||||||
% raftery_lewis [structure] containing the fields:
|
% raftery_lewis [structure] containing the fields:
|
||||||
% - M_burn [n_draws by 1] number of draws required for burn-in
|
% - M_burn [n_draws by 1] number of draws required for burn-in
|
||||||
|
@ -21,7 +21,7 @@ function [raftery_lewis] = raftery_lewis(runs,q,r,s)
|
||||||
% iterations due to dependence in chain
|
% iterations due to dependence in chain
|
||||||
% - N_min [scalar] # draws if the chain is white noise
|
% - N_min [scalar] # draws if the chain is white noise
|
||||||
% - N_total [n_draws by 1] nburn + nprec
|
% - N_total [n_draws by 1] nburn + nprec
|
||||||
%
|
%
|
||||||
|
|
||||||
% ---------------------------------------------------------------------
|
% ---------------------------------------------------------------------
|
||||||
% NOTES: Example values of q, r, s:
|
% NOTES: Example values of q, r, s:
|
||||||
|
@ -30,23 +30,23 @@ function [raftery_lewis] = raftery_lewis(runs,q,r,s)
|
||||||
%
|
%
|
||||||
% - The result is quite sensitive to r, being proportional to the
|
% - The result is quite sensitive to r, being proportional to the
|
||||||
% inverse of r^2.
|
% inverse of r^2.
|
||||||
% - For epsilon (closeness of probabilities to equilibrium values),
|
% - For epsilon (closeness of probabilities to equilibrium values),
|
||||||
% Raftery/Lewis use 0.001 and argue that the results
|
% Raftery/Lewis use 0.001 and argue that the results
|
||||||
% are quite robust to changes in this value
|
% are quite robust to changes in this value
|
||||||
%
|
%
|
||||||
% ---------------------------------------------------------------------
|
% ---------------------------------------------------------------------
|
||||||
% REFERENCES:
|
% REFERENCES:
|
||||||
% Raftery, Adrien E./Lewis, Steven (1992a): "How many iterations in the Gibbs sampler?"
|
% Raftery, Adrien E./Lewis, Steven (1992a): "How many iterations in the Gibbs sampler?"
|
||||||
% in: Bernardo/Berger/Dawid/Smith (eds.): Bayesian Statistics, Vol. 4, Clarendon Press: Oxford,
|
% in: Bernardo/Berger/Dawid/Smith (eds.): Bayesian Statistics, Vol. 4, Clarendon Press: Oxford,
|
||||||
% pp. 763-773.
|
% pp. 763-773.
|
||||||
% Raftery, Adrien E./Lewis, Steven (1992b): "Comment: One long run with diagnostics:
|
% Raftery, Adrien E./Lewis, Steven (1992b): "Comment: One long run with diagnostics:
|
||||||
% Implementation strategies for Markov chain Monte Carlo." Statistical Science,
|
% Implementation strategies for Markov chain Monte Carlo." Statistical Science,
|
||||||
% 7(4), pp. 493-497.
|
% 7(4), pp. 493-497.
|
||||||
%
|
%
|
||||||
% ----------------------------------------------------
|
% ----------------------------------------------------
|
||||||
|
|
||||||
% Copyright (C) 2016 Benjamin Born and Johannes Pfeifer
|
% Copyright (C) 2016 Benjamin Born and Johannes Pfeifer
|
||||||
% Copyright (C) 2016 Dynare Team
|
% Copyright (C) 2016-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -80,42 +80,42 @@ thinned_chain = zeros(n_runs,1);
|
||||||
Phi = norminv((s+1)/2); %note the missing ^{-1} at the Phi in equation top page 5, see RL (1995)
|
Phi = norminv((s+1)/2); %note the missing ^{-1} at the Phi in equation top page 5, see RL (1995)
|
||||||
raftery_lewis.N_min = fix(Phi^2*(1-q)*q/r^2+1);
|
raftery_lewis.N_min = fix(Phi^2*(1-q)*q/r^2+1);
|
||||||
|
|
||||||
for nv = 1:n_vars % big loop over variables
|
for nv = 1:n_vars % big loop over variables
|
||||||
if q > 0 && q < 1
|
if q > 0 && q < 1
|
||||||
work = (runs(:,nv) <= quantile(runs(:,nv),q));
|
work = (runs(:,nv) <= quantile(runs(:,nv),q));
|
||||||
else
|
else
|
||||||
error('Quantile must be between 0 and 1');
|
error('Quantile must be between 0 and 1');
|
||||||
end;
|
end
|
||||||
|
|
||||||
k_thin_current_var = 1;
|
k_thin_current_var = 1;
|
||||||
bic = 1;
|
bic = 1;
|
||||||
epss = 0.001;
|
epss = 0.001;
|
||||||
% Find thinning factor for which first-order Markov Chain is preferred to second-order one
|
% Find thinning factor for which first-order Markov Chain is preferred to second-order one
|
||||||
while(bic > 0)
|
while(bic > 0)
|
||||||
thinned_chain=work(1:k_thin_current_var:n_runs,1);
|
thinned_chain=work(1:k_thin_current_var:n_runs,1);
|
||||||
[g2, bic] = first_vs_second_order_MC_test(thinned_chain);
|
[g2, bic] = first_vs_second_order_MC_test(thinned_chain);
|
||||||
k_thin_current_var = k_thin_current_var+1;
|
k_thin_current_var = k_thin_current_var+1;
|
||||||
end;
|
end
|
||||||
|
|
||||||
k_thin_current_var = k_thin_current_var-1; %undo last step
|
k_thin_current_var = k_thin_current_var-1; %undo last step
|
||||||
|
|
||||||
%compute transition probabilities
|
%compute transition probabilities
|
||||||
transition_matrix = zeros(2,2);
|
transition_matrix = zeros(2,2);
|
||||||
for i1 = 2:size(thinned_chain,1)
|
for i1 = 2:size(thinned_chain,1)
|
||||||
transition_matrix(thinned_chain(i1-1)+1,thinned_chain(i1)+1) = transition_matrix(thinned_chain(i1-1)+1,thinned_chain(i1)+1)+1;
|
transition_matrix(thinned_chain(i1-1)+1,thinned_chain(i1)+1) = transition_matrix(thinned_chain(i1-1)+1,thinned_chain(i1)+1)+1;
|
||||||
end;
|
end
|
||||||
alpha = transition_matrix(1,2)/(transition_matrix(1,1)+transition_matrix(1,2)); %prob of going from 1 to 2
|
alpha = transition_matrix(1,2)/(transition_matrix(1,1)+transition_matrix(1,2)); %prob of going from 1 to 2
|
||||||
beta = transition_matrix(2,1)/(transition_matrix(2,1)+transition_matrix(2,2)); %prob of going from 2 to 1
|
beta = transition_matrix(2,1)/(transition_matrix(2,1)+transition_matrix(2,2)); %prob of going from 2 to 1
|
||||||
|
|
||||||
kmind=k_thin_current_var;
|
kmind=k_thin_current_var;
|
||||||
[g2, bic]=independence_chain_test(thinned_chain);
|
[g2, bic]=independence_chain_test(thinned_chain);
|
||||||
|
|
||||||
while(bic > 0)
|
while(bic > 0)
|
||||||
thinned_chain=work(1:kmind:n_runs,1);
|
thinned_chain=work(1:kmind:n_runs,1);
|
||||||
[g2, bic] = independence_chain_test(thinned_chain);
|
[g2, bic] = independence_chain_test(thinned_chain);
|
||||||
kmind = kmind+1;
|
kmind = kmind+1;
|
||||||
end;
|
end
|
||||||
|
|
||||||
m_star = log((alpha + beta)*epss/max(alpha,beta))/log(abs(1 - alpha - beta)); %equation bottom page 4
|
m_star = log((alpha + beta)*epss/max(alpha,beta))/log(abs(1 - alpha - beta)); %equation bottom page 4
|
||||||
raftery_lewis.M_burn(nv) = fix((m_star+1)*k_thin_current_var);
|
raftery_lewis.M_burn(nv) = fix((m_star+1)*k_thin_current_var);
|
||||||
n_star = (2 - (alpha + beta))*alpha*beta*(Phi^2)/((alpha + beta)^3 * r^2); %equation top page 5
|
n_star = (2 - (alpha + beta))*alpha*beta*(Phi^2)/((alpha + beta)^3 * r^2); %equation top page 5
|
||||||
|
@ -124,18 +124,18 @@ for nv = 1:n_vars % big loop over variables
|
||||||
raftery_lewis.k_ind(nv) = max(fix(raftery_lewis.I_stat(nv)+1),kmind);
|
raftery_lewis.k_ind(nv) = max(fix(raftery_lewis.I_stat(nv)+1),kmind);
|
||||||
raftery_lewis.k_thin(nv) = k_thin_current_var;
|
raftery_lewis.k_thin(nv) = k_thin_current_var;
|
||||||
raftery_lewis.N_total(nv)= raftery_lewis.M_burn(nv)+raftery_lewis.N_prec(nv);
|
raftery_lewis.N_total(nv)= raftery_lewis.M_burn(nv)+raftery_lewis.N_prec(nv);
|
||||||
end;
|
end
|
||||||
|
|
||||||
end
|
end
|
||||||
|
|
||||||
function [g2, bic] = first_vs_second_order_MC_test(d)
|
function [g2, bic] = first_vs_second_order_MC_test(d)
|
||||||
%conducts a test of first vs. second order Markov Chain via BIC criterion
|
%conducts a test of first vs. second order Markov Chain via BIC criterion
|
||||||
n_obs=size(d,1);
|
n_obs=size(d,1);
|
||||||
g2 = 0;
|
g2 = 0;
|
||||||
tran=zeros(2,2,2);
|
tran=zeros(2,2,2);
|
||||||
for t_iter=3:n_obs % count state transitions
|
for t_iter=3:n_obs % count state transitions
|
||||||
tran(d(t_iter-2,1)+1,d(t_iter-1,1)+1,d(t_iter,1)+1)=tran(d(t_iter-2,1)+1,d(t_iter-1,1)+1,d(t_iter,1)+1)+1;
|
tran(d(t_iter-2,1)+1,d(t_iter-1,1)+1,d(t_iter,1)+1)=tran(d(t_iter-2,1)+1,d(t_iter-1,1)+1,d(t_iter,1)+1)+1;
|
||||||
end;
|
end
|
||||||
% Compute the log likelihood ratio statistic for second-order MC vs first-order MC. G2 statistic of Bishop, Fienberg and Holland (1975)
|
% Compute the log likelihood ratio statistic for second-order MC vs first-order MC. G2 statistic of Bishop, Fienberg and Holland (1975)
|
||||||
for ind_1 = 1:2
|
for ind_1 = 1:2
|
||||||
for ind_2 = 1:2
|
for ind_2 = 1:2
|
||||||
|
@ -146,9 +146,9 @@ for ind_1 = 1:2
|
||||||
focus = tran(ind_1,ind_2,ind_3);
|
focus = tran(ind_1,ind_2,ind_3);
|
||||||
g2 = g2 + log(focus/fitted)*focus;
|
g2 = g2 + log(focus/fitted)*focus;
|
||||||
end
|
end
|
||||||
end; % end of for i3
|
end % end of for i3
|
||||||
end; % end of for i2
|
end % end of for i2
|
||||||
end; % end of for i1
|
end % end of for i1
|
||||||
g2 = g2*2;
|
g2 = g2*2;
|
||||||
bic = g2 - log(n_obs-2)*2;
|
bic = g2 - log(n_obs-2)*2;
|
||||||
|
|
||||||
|
@ -161,19 +161,19 @@ n_obs=size(d,1);
|
||||||
trans = zeros(2,2);
|
trans = zeros(2,2);
|
||||||
for ind_1 = 2:n_obs
|
for ind_1 = 2:n_obs
|
||||||
trans(d(ind_1-1)+1,d(ind_1)+1)=trans(d(ind_1-1)+1,d(ind_1)+1)+1;
|
trans(d(ind_1-1)+1,d(ind_1)+1)=trans(d(ind_1-1)+1,d(ind_1)+1)+1;
|
||||||
end;
|
end
|
||||||
dcm1 = n_obs - 1;
|
dcm1 = n_obs - 1;
|
||||||
g2 = 0;
|
g2 = 0;
|
||||||
% Compute the log likelihood ratio statistic for second-order MC vs first-order MC. G2 statistic of Bishop, Fienberg and Holland (1975)
|
% Compute the log likelihood ratio statistic for second-order MC vs first-order MC. G2 statistic of Bishop, Fienberg and Holland (1975)
|
||||||
for ind_1 = 1:2
|
for ind_1 = 1:2
|
||||||
for ind_2 = 1:2
|
for ind_2 = 1:2
|
||||||
if trans(ind_1,ind_2) ~= 0
|
if trans(ind_1,ind_2) ~= 0
|
||||||
fitted = ((trans(ind_1,1) + trans(ind_1,2))*(trans(1,ind_2) + trans(2,ind_2)))/dcm1;
|
fitted = ((trans(ind_1,1) + trans(ind_1,2))*(trans(1,ind_2) + trans(2,ind_2)))/dcm1;
|
||||||
focus = trans(ind_1,ind_2);
|
focus = trans(ind_1,ind_2);
|
||||||
g2 = g2 + log(focus/fitted)*focus;
|
g2 = g2 + log(focus/fitted)*focus;
|
||||||
end;
|
end
|
||||||
end;
|
end
|
||||||
end;
|
end
|
||||||
g2 = g2*2;
|
g2 = g2*2;
|
||||||
bic = g2 - log(dcm1);
|
bic = g2 - log(dcm1);
|
||||||
end
|
end
|
||||||
|
|
|
@ -19,7 +19,7 @@ function [info] = convertAimCodeToInfo(aimCode)
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% info [integer] Code to be used to print error in print_info.m
|
% info [integer] Code to be used to print error in print_info.m
|
||||||
|
|
||||||
% Copyright (C) 2011-2012 Dynare Team
|
% Copyright (C) 2011-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -55,7 +55,7 @@ switch aimCode
|
||||||
info = 161;
|
info = 161;
|
||||||
case 62
|
case 62
|
||||||
info = 162;
|
info = 162;
|
||||||
case 63;
|
case 63
|
||||||
info = 163;
|
info = 163;
|
||||||
case 64
|
case 64
|
||||||
info = 164;
|
info = 164;
|
||||||
|
|
|
@ -13,7 +13,7 @@ function oo_ = convert_dyn_45_to_44(M_, options_, oo_,bayestopt_)
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2015-2016 Dynare Team
|
% Copyright (C) 2015-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -40,7 +40,7 @@ if isfield(oo_,'PointForecast')
|
||||||
oo_.MeanForecast.(moment_names{moment_iter}).(var_names{var_iter})=...
|
oo_.MeanForecast.(moment_names{moment_iter}).(var_names{var_iter})=...
|
||||||
[oo_.SmoothedVariables.(moment_names{moment_iter}).(var_names{var_iter})(:,end)*ones(M_.maximum_endo_lag,1) oo_.MeanForecast.(moment_names{moment_iter}).(var_names{var_iter})];
|
[oo_.SmoothedVariables.(moment_names{moment_iter}).(var_names{var_iter})(:,end)*ones(M_.maximum_endo_lag,1) oo_.MeanForecast.(moment_names{moment_iter}).(var_names{var_iter})];
|
||||||
oo_.PointForecast.(moment_names{moment_iter}).(var_names{var_iter})=...
|
oo_.PointForecast.(moment_names{moment_iter}).(var_names{var_iter})=...
|
||||||
[oo_.SmoothedVariables.(moment_names{moment_iter}).(var_names{var_iter})(:,end)*ones(M_.maximum_endo_lag,1) oo_.PointForecast.(moment_names{moment_iter}).(var_names{var_iter})];
|
[oo_.SmoothedVariables.(moment_names{moment_iter}).(var_names{var_iter})(:,end)*ones(M_.maximum_endo_lag,1) oo_.PointForecast.(moment_names{moment_iter}).(var_names{var_iter})];
|
||||||
else
|
else
|
||||||
oo_.MeanForecast.(moment_names{moment_iter}).(var_names{var_iter})=...
|
oo_.MeanForecast.(moment_names{moment_iter}).(var_names{var_iter})=...
|
||||||
[oo_.SmoothedVariables.(moment_names{moment_iter}).(var_names{var_iter})(end)*ones(M_.maximum_endo_lag,1); oo_.MeanForecast.(moment_names{moment_iter}).(var_names{var_iter})];
|
[oo_.SmoothedVariables.(moment_names{moment_iter}).(var_names{var_iter})(end)*ones(M_.maximum_endo_lag,1); oo_.MeanForecast.(moment_names{moment_iter}).(var_names{var_iter})];
|
||||||
|
@ -106,7 +106,7 @@ if isfield(oo_,'UpdatedVariables')
|
||||||
for ii=1:length(names)
|
for ii=1:length(names)
|
||||||
%make sure Bayesian fields are not affected
|
%make sure Bayesian fields are not affected
|
||||||
if ~strcmp(names{ii},'Mean') && ~strcmp(names{ii},'Median') && ~strcmp(names{ii},'deciles') ...
|
if ~strcmp(names{ii},'Mean') && ~strcmp(names{ii},'Median') && ~strcmp(names{ii},'deciles') ...
|
||||||
&& ~strcmp(names{ii},'Var') && ~strcmp(names{ii},'HPDinf') && ~strcmp(names{ii},'HPDsup')
|
&& ~strcmp(names{ii},'Var') && ~strcmp(names{ii},'HPDinf') && ~strcmp(names{ii},'HPDsup')
|
||||||
current_var_index=find(strmatch(names{ii},deblank(M_.endo_names),'exact'));
|
current_var_index=find(strmatch(names{ii},deblank(M_.endo_names),'exact'));
|
||||||
if options_.loglinear == 1 %logged steady state must be used
|
if options_.loglinear == 1 %logged steady state must be used
|
||||||
constant_current_variable=log(oo_.dr.ys(current_var_index));
|
constant_current_variable=log(oo_.dr.ys(current_var_index));
|
||||||
|
@ -127,7 +127,7 @@ if isfield(oo_,'FilteredVariables')
|
||||||
for ii=1:length(names)
|
for ii=1:length(names)
|
||||||
%make sure Bayesian fields are not affected
|
%make sure Bayesian fields are not affected
|
||||||
if ~strcmp(names{ii},'Mean') && ~strcmp(names{ii},'Median') && ~strcmp(names{ii},'deciles') ...
|
if ~strcmp(names{ii},'Mean') && ~strcmp(names{ii},'Median') && ~strcmp(names{ii},'deciles') ...
|
||||||
&& ~strcmp(names{ii},'Var') && ~strcmp(names{ii},'HPDinf') && ~strcmp(names{ii},'HPDsup')
|
&& ~strcmp(names{ii},'Var') && ~strcmp(names{ii},'HPDinf') && ~strcmp(names{ii},'HPDsup')
|
||||||
current_var_index=find(strmatch(names{ii},deblank(M_.endo_names),'exact'));
|
current_var_index=find(strmatch(names{ii},deblank(M_.endo_names),'exact'));
|
||||||
if options_.loglinear == 1 %logged steady state must be used
|
if options_.loglinear == 1 %logged steady state must be used
|
||||||
constant_current_variable=log(oo_.dr.ys(current_var_index));
|
constant_current_variable=log(oo_.dr.ys(current_var_index));
|
||||||
|
|
|
@ -15,7 +15,7 @@ function oo_ = convert_oo_(M_, options_, oo_, from_ver, to_ver)
|
||||||
% SPECIAL REQUIREMENTS
|
% SPECIAL REQUIREMENTS
|
||||||
% none
|
% none
|
||||||
|
|
||||||
% Copyright (C) 2015 Dynare Team
|
% Copyright (C) 2015-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -63,7 +63,7 @@ else
|
||||||
end
|
end
|
||||||
|
|
||||||
if strcmp(from_ver, to_ver)
|
if strcmp(from_ver, to_ver)
|
||||||
return;
|
return
|
||||||
end
|
end
|
||||||
|
|
||||||
if ver_greater_than(to_ver, from_ver)
|
if ver_greater_than(to_ver, from_ver)
|
||||||
|
|
|
@ -2,7 +2,7 @@ function oo_ = correlation_mc_analysis(SampleSize,type,dname,fname,vartan,nvar,v
|
||||||
% This function analyses the (posterior or prior) distribution of the
|
% This function analyses the (posterior or prior) distribution of the
|
||||||
% endogenous variables correlation function.
|
% endogenous variables correlation function.
|
||||||
|
|
||||||
% Copyright (C) 2008-2013 Dynare Team
|
% Copyright (C) 2008-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -137,7 +137,7 @@ end
|
||||||
|
|
||||||
function oo_ = fill_output_structure(var1,var2,type,oo_,moment,lag,result)
|
function oo_ = fill_output_structure(var1,var2,type,oo_,moment,lag,result)
|
||||||
switch moment
|
switch moment
|
||||||
case {'Mean','Median','Variance','HPDinf','HPDsup'}
|
case {'Mean','Median','Variance','HPDinf','HPDsup'}
|
||||||
oo_.([type, 'TheoreticalMoments']).dsge.correlation.(moment).(var1).(var2)(lag,1) = result;
|
oo_.([type, 'TheoreticalMoments']).dsge.correlation.(moment).(var1).(var2)(lag,1) = result;
|
||||||
case {'deciles','density'}
|
case {'deciles','density'}
|
||||||
oo_.([type, 'TheoreticalMoments']).dsge.correlation.(moment).(var1).(var2)(lag,1) = {result};
|
oo_.([type, 'TheoreticalMoments']).dsge.correlation.(moment).(var1).(var2)(lag,1) = {result};
|
||||||
|
|
|
@ -1,4 +1,4 @@
|
||||||
function [co, b, yhat] = cosn(H);
|
function [co, b, yhat] = cosn(H)
|
||||||
|
|
||||||
% function co = cosn(H);
|
% function co = cosn(H);
|
||||||
% computes the cosine of the angle between the H(:,1) and its
|
% computes the cosine of the angle between the H(:,1) and its
|
||||||
|
@ -7,7 +7,7 @@ function [co, b, yhat] = cosn(H);
|
||||||
% Not the same as multiple correlation coefficient since the means are not
|
% Not the same as multiple correlation coefficient since the means are not
|
||||||
% zero
|
% zero
|
||||||
%
|
%
|
||||||
% Copyright (C) 2008-2012 Dynare Team
|
% Copyright (C) 2008-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -28,15 +28,12 @@ y = H(:,1);
|
||||||
X = H(:,2:end);
|
X = H(:,2:end);
|
||||||
|
|
||||||
b=(X\y);
|
b=(X\y);
|
||||||
if any(isnan(b)) || any(isinf(b)),
|
if any(isnan(b)) || any(isinf(b))
|
||||||
b=0;
|
b=0;
|
||||||
end
|
end
|
||||||
yhat = X*b;
|
yhat = X*b;
|
||||||
if rank(yhat),
|
if rank(yhat)
|
||||||
co = abs(y'*yhat/sqrt((y'*y)*(yhat'*yhat)));
|
co = abs(y'*yhat/sqrt((y'*y)*(yhat'*yhat)));
|
||||||
else
|
else
|
||||||
co=0;
|
co=0;
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
|
@ -1,7 +1,7 @@
|
||||||
function oo_ = covariance_mc_analysis(NumberOfSimulations,type,dname,fname,vartan,nvar,var1,var2,mh_conf_sig,oo_,options_)
|
function oo_ = covariance_mc_analysis(NumberOfSimulations,type,dname,fname,vartan,nvar,var1,var2,mh_conf_sig,oo_,options_)
|
||||||
% This function analyses the (posterior or prior) distribution of the
|
% This function analyses the (posterior or prior) distribution of the
|
||||||
% endogenous variables' covariance matrix.
|
% endogenous variables' covariance matrix.
|
||||||
%
|
%
|
||||||
% INPUTS
|
% INPUTS
|
||||||
% NumberOfSimulations [integer] scalar, number of simulations.
|
% NumberOfSimulations [integer] scalar, number of simulations.
|
||||||
% type [string] 'prior' or 'posterior'
|
% type [string] 'prior' or 'posterior'
|
||||||
|
@ -14,12 +14,12 @@ function oo_ = covariance_mc_analysis(NumberOfSimulations,type,dname,fname,varta
|
||||||
% mh_conf_sig [double] 2 by 1 vector with upper
|
% mh_conf_sig [double] 2 by 1 vector with upper
|
||||||
% and lower bound of HPD intervals
|
% and lower bound of HPD intervals
|
||||||
% oo_ [structure] Dynare structure where the results are saved.
|
% oo_ [structure] Dynare structure where the results are saved.
|
||||||
% options_ [structure] Dynare options structure
|
% options_ [structure] Dynare options structure
|
||||||
%
|
%
|
||||||
% OUTPUTS
|
% OUTPUTS
|
||||||
% oo_ [structure] Dynare structure where the results are saved.
|
% oo_ [structure] Dynare structure where the results are saved.
|
||||||
|
|
||||||
% Copyright (C) 2008-2015 Dynare Team
|
% Copyright (C) 2008-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -64,7 +64,7 @@ var1=deblank(var1);
|
||||||
var2=deblank(var2);
|
var2=deblank(var2);
|
||||||
|
|
||||||
if isfield(oo_,[ TYPE 'TheoreticalMoments'])
|
if isfield(oo_,[ TYPE 'TheoreticalMoments'])
|
||||||
temporary_structure = oo_.([TYPE, 'TheoreticalMoments']);
|
temporary_structure = oo_.([TYPE, 'TheoreticalMoments']);
|
||||||
if isfield(temporary_structure,'dsge')
|
if isfield(temporary_structure,'dsge')
|
||||||
temporary_structure = oo_.([TYPE, 'TheoreticalMoments']).dsge;
|
temporary_structure = oo_.([TYPE, 'TheoreticalMoments']).dsge;
|
||||||
if isfield(temporary_structure,'covariance')
|
if isfield(temporary_structure,'covariance')
|
||||||
|
@ -104,7 +104,7 @@ for file = 1:length(ListOfFiles)
|
||||||
temp=Covariance_matrix(:,cov_pos)./(sqrt(Covariance_matrix(:,var_pos_1)).*sqrt(Covariance_matrix(:,var_pos_2)));
|
temp=Covariance_matrix(:,cov_pos)./(sqrt(Covariance_matrix(:,var_pos_1)).*sqrt(Covariance_matrix(:,var_pos_2)));
|
||||||
temp(Covariance_matrix(:,cov_pos)==0)=0; %filter out 0 correlations that would result in 0/0
|
temp(Covariance_matrix(:,cov_pos)==0)=0; %filter out 0 correlations that would result in 0/0
|
||||||
tmp_corr_mat(i1:i2)=temp;
|
tmp_corr_mat(i1:i2)=temp;
|
||||||
end
|
end
|
||||||
i1 = i2+1;
|
i1 = i2+1;
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
|
@ -12,7 +12,7 @@ function [x,rc] = csolve(FUN,x,gradfun,crit,itmax,varargin)
|
||||||
% the equation is solved.
|
% the equation is solved.
|
||||||
% itmax: the solver stops when this number of iterations is reached, with rc=4
|
% itmax: the solver stops when this number of iterations is reached, with rc=4
|
||||||
% varargin: in this position the user can place any number of additional arguments, all
|
% varargin: in this position the user can place any number of additional arguments, all
|
||||||
% of which are passed on to FUN and gradfun (when it is non-empty) as a list of
|
% of which are passed on to FUN and gradfun (when it is non-empty) as a list of
|
||||||
% arguments following x.
|
% arguments following x.
|
||||||
% rc: 0 means normal solution, 1 and 3 mean no solution despite extremely fine adjustments
|
% rc: 0 means normal solution, 1 and 3 mean no solution despite extremely fine adjustments
|
||||||
% in step length (very likely a numerical problem, or a discontinuity). 4 means itmax
|
% in step length (very likely a numerical problem, or a discontinuity). 4 means itmax
|
||||||
|
@ -22,7 +22,7 @@ function [x,rc] = csolve(FUN,x,gradfun,crit,itmax,varargin)
|
||||||
% http://sims.princeton.edu/yftp/optimize/mfiles/csolve.m
|
% http://sims.princeton.edu/yftp/optimize/mfiles/csolve.m
|
||||||
|
|
||||||
% Copyright (C) 1993-2007 Christopher Sims
|
% Copyright (C) 1993-2007 Christopher Sims
|
||||||
% Copyright (C) 2007-2011 Dynare Team
|
% Copyright (C) 2007-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -60,7 +60,7 @@ if isempty(varargin)
|
||||||
f0=feval(FUN,x);
|
f0=feval(FUN,x);
|
||||||
else
|
else
|
||||||
f0=feval(FUN,x,varargin{:});
|
f0=feval(FUN,x,varargin{:});
|
||||||
end
|
end
|
||||||
af0=sum(abs(f0));
|
af0=sum(abs(f0));
|
||||||
af00=af0;
|
af00=af0;
|
||||||
itct=0;
|
itct=0;
|
||||||
|
@ -122,7 +122,7 @@ while ~done
|
||||||
factor=factor^.6;
|
factor=factor^.6;
|
||||||
shrink=1;
|
shrink=1;
|
||||||
end
|
end
|
||||||
if abs(lambda*(1-factor))*dxSize > .1*delta;
|
if abs(lambda*(1-factor))*dxSize > .1*delta
|
||||||
lambda = factor*lambda;
|
lambda = factor*lambda;
|
||||||
elseif (lambda > 0) && (factor==.6) %i.e., we've only been shrinking
|
elseif (lambda > 0) && (factor==.6) %i.e., we've only been shrinking
|
||||||
lambda=-.3;
|
lambda=-.3;
|
||||||
|
@ -162,7 +162,7 @@ while ~done
|
||||||
if itct >= itmax
|
if itct >= itmax
|
||||||
done=1;
|
done=1;
|
||||||
rc=4;
|
rc=4;
|
||||||
elseif af0<crit;
|
elseif af0<crit
|
||||||
done=1;
|
done=1;
|
||||||
rc=0;
|
rc=0;
|
||||||
end
|
end
|
||||||
|
|
|
@ -29,12 +29,12 @@ function [nodes, weights] = cubature_with_gaussian_weight(d,n,method)
|
||||||
%! The routine returns nodes and associated weights to compute a multivariate integral of the form:
|
%! The routine returns nodes and associated weights to compute a multivariate integral of the form:
|
||||||
%!
|
%!
|
||||||
%! \int_D f(x)*\exp(-<x,x>) dx
|
%! \int_D f(x)*\exp(-<x,x>) dx
|
||||||
%!
|
%!
|
||||||
%!
|
%!
|
||||||
%! @end deftypefn
|
%! @end deftypefn
|
||||||
%@eod:
|
%@eod:
|
||||||
|
|
||||||
% Copyright (C) 2012-2013 Dynare Team
|
% Copyright (C) 2012-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -90,7 +90,7 @@ if strcmp(method,'Stroud') && isequal(n,5)
|
||||||
nodes = zeros(d,m);
|
nodes = zeros(d,m);
|
||||||
weights = zeros(m,1);
|
weights = zeros(m,1);
|
||||||
% Set the weight for the first node (0)
|
% Set the weight for the first node (0)
|
||||||
weights(1) = A;
|
weights(1) = A;
|
||||||
skip = 1;
|
skip = 1;
|
||||||
% Set the remaining nodes and associated weights.
|
% Set the remaining nodes and associated weights.
|
||||||
nodes(:,skip+(1:d)) = r*eye(d);
|
nodes(:,skip+(1:d)) = r*eye(d);
|
||||||
|
@ -130,7 +130,7 @@ m(:,4) = -m(:,1);
|
||||||
%@test:1
|
%@test:1
|
||||||
%$ % Set problem
|
%$ % Set problem
|
||||||
%$ d = 4;
|
%$ d = 4;
|
||||||
%$
|
%$
|
||||||
%$ t = zeros(5,1);
|
%$ t = zeros(5,1);
|
||||||
%$
|
%$
|
||||||
%$ % Call the tested routine
|
%$ % Call the tested routine
|
||||||
|
@ -289,7 +289,7 @@ m(:,4) = -m(:,1);
|
||||||
%@test:5
|
%@test:5
|
||||||
%$ % Set problem
|
%$ % Set problem
|
||||||
%$ d = 5;
|
%$ d = 5;
|
||||||
%$
|
%$
|
||||||
%$ t = zeros(6,1);
|
%$ t = zeros(6,1);
|
||||||
%$
|
%$
|
||||||
%$ % Call the tested routine
|
%$ % Call the tested routine
|
||||||
|
@ -333,7 +333,7 @@ m(:,4) = -m(:,1);
|
||||||
%@test:6
|
%@test:6
|
||||||
%$ % Set problem
|
%$ % Set problem
|
||||||
%$ d = 3;
|
%$ d = 3;
|
||||||
%$
|
%$
|
||||||
%$ t = zeros(4,1);
|
%$ t = zeros(4,1);
|
||||||
%$
|
%$
|
||||||
%$ % Call the tested routine
|
%$ % Call the tested routine
|
||||||
|
|
|
@ -43,7 +43,7 @@ function [X, info] = cycle_reduction(A0, A1, A2, cvg_tol, ch) % --*-- Unitary te
|
||||||
%! @end deftypefn
|
%! @end deftypefn
|
||||||
%@eod:
|
%@eod:
|
||||||
|
|
||||||
% Copyright (C) 2013 Dynare Team
|
% Copyright (C) 2013-2017 Dynare Team
|
||||||
%
|
%
|
||||||
% This file is part of Dynare.
|
% This file is part of Dynare.
|
||||||
%
|
%
|
||||||
|
@ -76,7 +76,7 @@ id0 = 1:n;
|
||||||
id2 = id0+n;
|
id2 = id0+n;
|
||||||
|
|
||||||
cont = 1;
|
cont = 1;
|
||||||
while cont
|
while cont
|
||||||
tmp = ([A0; A2]/A1)*[A0 A2];
|
tmp = ([A0; A2]/A1)*[A0 A2];
|
||||||
A1 = A1 - tmp(id0,id2) - tmp(id2,id0);
|
A1 = A1 - tmp(id0,id2) - tmp(id2,id0);
|
||||||
A0 = -tmp(id0,id0);
|
A0 = -tmp(id0,id0);
|
||||||
|
@ -97,7 +97,7 @@ while cont
|
||||||
info(2) = log(norm(A1,1));
|
info(2) = log(norm(A1,1));
|
||||||
end
|
end
|
||||||
return
|
return
|
||||||
end
|
end
|
||||||
it = it + 1;
|
it = it + 1;
|
||||||
end
|
end
|
||||||
|
|
||||||
|
|
Some files were not shown because too many files have changed in this diff Show More
Loading…
Reference in New Issue