NEWS 73.8 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
Announcement for Dynare 4.5.2 (on 2017-10-19)
=============================================

We are pleased to announce the release of Dynare 4.5.2.

This is a bugfix release.

The Windows packages are already available for download at:

 http://www.dynare.org/download/dynare-stable

The Mac and GNU/Linux packages (for Debian and Ubuntu LTS) should follow soon.

This release is compatible with MATLAB versions 7.3 (R2006b) to 9.3 (R2017b)
and with GNU Octave versions 4.2.

Here is a list of the problems identified in version 4.5.1 and that have been
fixed in version 4.5.2:


 - Fixed bug in perfect foresight solver:

   + If expected shocks were declared after the terminal period, as specified
   by the `periods` option, Dynare was crashing.

   + Models declared with the `linear` option were crashing if exogenous
   variables were present with a lead or lag.

 - After ML or Bayesian estimation when the smoother option or `mh_replic=0`
   were not specified, not all smoothed measurement errors were displayed.

 - Fixed error in reference manual about the `conditional_forecasts` command.

 - Fixed smoother behaviour, provide informative error instead of crashing when
   model cannot be solved.

 - The `nopathchange` preprocessor option was always triggered, regardless of
   whether it was passed or not.

 - When `ramsey_policy` is used, allow state variables to be set in `histval`
   block.

 - `histval` erroneously accepted leads, leading to cryptic crashes.

 - The prior MC draws from previous runs were not deleted, potentially
   resulting in loading stale files.

 - `estim_params_` was being declared `global` more than once.

 - Fixed crashes happening when simulating linear models with order>1.

 - Make empirical moments independent of `simul_replic`, as stated in the
   reference manual, by outputting moments computed with the first simulated
   sample.

 - The `prior_function` required a preceding `estimation`-command to properly
   set up the prior.

 - If the mode for a parameter was at exactly 0, `mode_check` was crashing.

 - Fixed `get_posterior_parameters`-routine which should not do more than
   getting parameters. As a consequecence, the `shock_decomposition`-command
   did not correctly set the `parameter_set` for use in subsequent function
   calls if shocks are correlated or measurement error is present.

 - Fixed bug in Ramsey problem with constraints both on a policy instrument and
   another variable. Note that the constraint on a variable that is not an
   instrument of the Ramsey problem must be written with an equation tag in the
   model block.
 
 - Fixed bug in Ramsey problem with constraints on policy instrument.

 - Fixed crash with optimizer 5 when not used with DSGE model at order 1.

 - Fixed mex file used for third order approximation (was crashing on
   Matlab/Windows 7).



80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
Announcement for Dynare 4.5.1 (on 2017-08-24)
=============================================

We are pleased to announce the release of Dynare 4.5.1.

This is a bugfix release.

The Windows packages are already available for download at:

 http://www.dynare.org/download/dynare-stable

The Mac and GNU/Linux packages (for Debian and Ubuntu LTS) should follow soon.

This release is compatible with MATLAB versions 7.3 (R2006b) to 9.2 (R2017a)
and with GNU Octave versions 4.2.

Here is a list of the problems identified in version 4.5.0 and that have been
fixed in version 4.5.1:


 - Fixed out of memory issue with simpsa optimization algorithm.

 - Added missing plots for measurement errors with `generate_trace_plot`
   command.

 - Posterior moments after MCMC for very big models were not correctly computed
   and their plotting might crash Dynare.

 - Results of the posterior conditional variance decomposition after MCMC were
   not correctly computed.

 - Options `use_shock_groups` and `colormap` of the `shock_decomposition`
   command were not working.

 - Added a clean error message if sensitivity toolbox is used with recursive
   estimation.

 - Computation of posterior filtered variables was crashing in models with only
   one variable.

 - Fixed various typos and errors in the reference manual.



124
Announcement for Dynare 4.5.0 (on 2017-06-11)
Stéphane Adjemian's avatar
Stéphane Adjemian committed
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
=============================================

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:


 - Ramsey policy

147
148
149
150
   + 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,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
151

152
153
   + `ramsey_policy` accepts multipliers in its variable list and
     displays results for them.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
154
155
156
157


 - Perfect foresight models

158
159
160
161
162
163
164
165
   + 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`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
166

167
168
169
   + 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,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
170

171
172
   + New option `stack_solve_algo=7` that allows specifying a
     `solve_algo` solver for solving the model,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
173

174
175
   + New option `solve_algo` that allows specifying a solver for
     solving the model when using `stack_solve_algo=7`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
176

177
178
   + New option `lmmcp` that solves the model via a Levenberg-Marquardt
     mixed complementarity problem (LMMCP) solver,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
179

180
181
   + New option `robust_lin_solve` that triggers the use of a robust
     linear solver for the default `solve_algo=4`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
182

183
184
   + New options `tolf` and `tolx` to control termination criteria of
     solvers,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
185

186
   + New option `endogenous_terminal_period` to `simul`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
187

188
189
   + Added the possibility to set the initial condition of the
     (stochastic) extended path simulations with the histval block.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
190
191
192
193


 - Optimal simple rules

194
   + Saves the optimal value of parameters to `oo_.osr.optim_params`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
195

196
197
   + New block `osr_params_bounds` allows specifying bounds for the
     estimated parameters,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
198

199
200
   + New option `opt_algo` allows selecting different optimizers while
     the new option `optim` allows specifying the optimizer options,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
201

202
203
204
   + 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`.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
205
206
207
208


 - Forecasts and Smoothing

209
210
   + The smoother and forecasts take uncertainty about trends and means
     into account,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
211

212
213
   + Forecasts accounting for measurement error are now saved in fields
     of the form `HPDinf_ME` and `HPDsup_ME`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
214

215
216
   + New fields `oo_.Smoother.Trend` and `oo_.Smoother.Constant` that
     save the trend and constant parts of the smoothed variables,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
217

218
219
   + new field `oo_.Smoother.TrendCoeffs` that stores the trend
     coefficients.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
220

221
222
   + Rolling window forecasts allowed in `estimation` command by
     passing a vector to `first_obs`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
223

224
225
   + The `calib_smoother` command now accepts the `loglinear`,
     `prefilter`, `first_obs` and `filter_decomposition` options.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
226
227
228
229


 - Estimation

230
231
232
   + New options: `logdata`, `consider_all_endogenous`,
     `consider_only_observed`, `posterior_max_subsample_draws`,
     `mh_conf_sig`, `diffuse_kalman_tol`, `dirname`, `nodecomposition`
Stéphane Adjemian's avatar
Stéphane Adjemian committed
233

234
   + `load_mh_file` and `mh_recover` now try to load chain's proposal density,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
235

236
237
238
   + New option `load_results_after_load_mh` that allows loading some
     posterior results from a previous run if no new MCMC draws are
     added,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
239

240
241
242
   + New option `posterior_nograph` that suppresses the generation of
     graphs associated with Bayesian IRFs, posterior smoothed objects,
     and posterior forecasts,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
243

244
245
   + Saves the posterior density at the mode in
     `oo_.posterior.optimization.log_density`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
246

247
248
   + The `filter_covariance` option now also works with posterior
     sampling like Metropolis-Hastings,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
249

250
251
   + New option `no_posterior_kernel_density` to suppress computation
     of kernel density of posterior objects,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
252

253
254
   + Recursive estimation and forecasting now provides the individual
     `oo_` structures for each sample in `oo_recursive_`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
255

256
   + The `trace_plot` command can now plot the posterior density,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
257

258
259
   + New command `generate_trace_plots` allows generating all trace
     plots for one chain,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
260

261
262
263
   + New commands `prior_function` and `posterior_function` that
     execute a user-defined function on parameter draws from the
     prior/posterior distribution,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
264

265
266
   + New option `huge_number` for replacement of infinite bounds with
     large number during `mode_compute`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
267

268
269
270
271
272
273
   + 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),
Stéphane Adjemian's avatar
Stéphane Adjemian committed
274

275
276
277
278
   + 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`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
279

280
281
   + New option `raftery_lewis_diagnostics` that computes Raftery/Lewis
     (1992) convergence diagnostics,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
282

283
284
   + New option `fast_kalman_filter` that provides fast Kalman filter
     using Chandrasekhar recursions as described in Ed Herbst (2015),
Stéphane Adjemian's avatar
Stéphane Adjemian committed
285

286
287
   + The `dsge_var` option now saves results at the posterior mode into
     `oo_.dsge_var`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
288

289
290
   + New option `smoothed_state_uncertainty` to provide the uncertainty
     estimate for the smoothed state estimate from the Kalman smoother,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
291

292
   + New prior density: generalized Weibull distribution,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
293

294
295
   + Option `mh_recover` now allows continuing a crashed chain at the
     last save mh-file,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
296

297
298
299
   + New option `nonlinear_filter_initialization` for the
     `estimation` command. Controls the initial covariance matrix
     of the state variables in nonlinear filters.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
300

301
302
303
   + The `conditional_variance_decomposition` option now displays
     output and stores it as a LaTeX-table when the `TeX` option is
     invoked,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
304

305
306
   + The `use_calibration` to `estimated_params_init` now also works
     with ML,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
307

308
   + Improved initial estimation checks.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
309
310
311
312


 - Steady state

313
314
315
   + The default solver for finding the steady state is now a
     trust-region solver (can be triggered explicitly with option
     `solve_algo=4`),
Stéphane Adjemian's avatar
Stéphane Adjemian committed
316

317
318
   + New options `tolf` and `tolx` to control termination criteria of
     solver,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
319

320
321
   + The debugging mode now provides the termination values in steady
     state finding.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
322
323
324
325


 - Stochastic simulations

326
   + New options `nodecomposition`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
327

328
329
   + New option `bandpass_filter` to compute bandpass-filtered
     theoretical and simulated moments,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
330

331
332
   + New option `one_sided_hp_filter` to compute one-sided HP-filtered
     simulated moments,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
333

334
335
   + `stoch_simul` displays a simulated variance decomposition when
     simulated moments are requested,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
336

337
338
   + `stoch_simul` saves skewness and kurtosis into respective fields
     of `oo_` when simulated moments have been requested,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
339

340
341
   + `stoch_simul` saves the unconditional variance decomposition in
     `oo_.variance_decomposition`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
342

343
344
   + New option `dr_display_tol` that governs omission of small terms
     in display of decision rules,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
345

346
347
   + The `stoch_simul` command now prints the displayed tables as LaTeX
     code when the new `TeX` option is enabled,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
348

349
350
   + The `loglinear` option now works with lagged and leaded exogenous
     variables like news shocks,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
351

352
353
   + New option `spectral_density` that allows displaying the spectral
     density of (filtered) endogenous variables,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
354

355
356
   + New option `contemporaneous_correlation` that allows saving
     contemporaneous correlations in addition to the covariances.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
357
358
359
360


 - Identification

361
   + New options `diffuse_filter` and `prior_trunc`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
362

363
364
   + The `identification` command now supports correlations via
     simulated moments,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
365
366
367
368


 - Sensitivity analysis

369
   + New blocks `irf_calibration` and `moment_calibration`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
370

371
   + Outputs LaTeX tables if the new `TeX` option is used,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
372

373
   + New option `relative_irf` to `irf_calibration` block.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
374
375
376
377


 - Conditional forecast

378
379
   + Command `conditional_forecast` now takes into account `histval`
     block if present.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
380
381
382
383


 - Shock decomposition

384
385
   + New option `colormap` to `shocks_decomposition` for controlling
     the color map used in the shocks decomposition graphs,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
386

387
   + `shocks_decomposition` now accepts the `nograph` option,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
388

389
390
   + New command `realtime_shock_decomposition` that for each period `T= [presample,...,nobs]`
     allows computing the:
Stéphane Adjemian's avatar
Stéphane Adjemian committed
391

392
     * realtime historical shock decomposition `Y(t|T)`, i.e. without observing data in `[T+1,...,nobs]`
Stéphane Adjemian's avatar
Stéphane Adjemian committed
393

394
     * forecast shock decomposition `Y(T+k|T)`
Stéphane Adjemian's avatar
Stéphane Adjemian committed
395

396
397
398
399
400
401
402
403
404
     * 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.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
405
406
407
408


 - Macroprocessor

409
   + Can now pass a macro-variable to the `@#include` macro directive,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
410

411
412
413
   + 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`.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
414
415
416
417


 - Command line

418
419
   + New option `onlyclearglobals` (do not clear JIT compiled functions
     with recent versions of Matlab),
Stéphane Adjemian's avatar
Stéphane Adjemian committed
420

421
422
   + New option `minimal_workspace` to use fewer variables in the
     current workspace,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
423

424
425
426
   + New option `params_derivs_order` allows limiting the order of the
     derivatives with respect to the parameters that are calculated by
     the preprocessor,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
427

428
429
   + New command line option `mingw` to support the MinGW-w64 C/C++
     Compiler from TDM-GCC for `use_dll`.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
430
431
432
433


 - dates/dseries/reporting classes

434
   + New methods `abs`, `cumprod` and `chain`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
435

436
   + New option `tableRowIndent` to `addTable`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
437

438
439
   + Reporting system revamped and made more efficient, dependency on
     matlab2tikz has been dropped.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
440
441
442
443


 - Optimization algorithms

444
445
   + `mode_compute=2` Uses the simulated annealing as described by
     Corana et al. (1987),
Stéphane Adjemian's avatar
Stéphane Adjemian committed
446

447
448
   + `mode_compute=101` Uses SOLVEOPT as described by Kuntsevich and
     Kappel (1997),
Stéphane Adjemian's avatar
Stéphane Adjemian committed
449

450
451
   + `mode_compute=102` Uses `simulannealbnd` from Matlab's Global
     Optimization Toolbox (if available),
Stéphane Adjemian's avatar
Stéphane Adjemian committed
452

453
454
   + New option `silent_optimizer` to shut off output from mode
     computing/optimization,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
455

456
457
   + New options `verbosity` and `SaveFiles` to control output and
     saving of files during mode computing/optimization.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
458
459
460
461


 - LaTeX output

462
   + New command `write_latex_original_model`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
463

464
465
466
   + New option `write_equation_tags` to `write_latex_dynamic_model`
     that allows printing the specified equation tags to the generate
     LaTeX code,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
467

468
469
   + New command `write_latex_parameter_table` that writes the names and
     values of model parameters to a LaTeX table,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
470

471
472
   + New command `write_latex_prior_table` that writes the descriptive
     statistics about the prior distribution to a LaTeX table,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
473

474
475
   + New command `collect_latex_files` that creates one compilable LaTeX
     file containing all TeX-output.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
476
477
478
479


 - Misc.

480
   + Provides 64bit preprocessor,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
481

482
483
   + Introduces new path management to avoid conflicts with other
     toolboxes,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
484

485
   + Full compatibility with Matlab 2014b's new graphic interface,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
486

487
488
   + When using `model(linear)`, Dynare automatically checks
     whether the model is truly linear,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
489

490
491
   + `usedll`, the `msvc` option now supports `normcdf`, `acosh`,
     `asinh`, and `atanh`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
492

493
494
495
   + New parallel option `NumberOfThreadsPerJob` for Windows nodes that
     sets the number of threads assigned to each remote MATLAB/Octave
     run,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
496

497
498
   + Improved numerical performance of
     `schur_statespace_transformation` for very large models,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
499

500
   + The `all_values_required` option now also works with `histval`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
501

502
   + Add missing `horizon` option to `ms_forecast`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
503

504
505
506
507
   + 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`.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
508
509
510
511
512
513
514
515



* Bugs and problems identified in version 4.4.3 and that have been fixed in version 4.5.0:


 - BVAR models

516
517
   + `bvar_irf` could display IRFs in an unreadable way when they moved from
     negative to positive values,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
518

519
520
   + In contrast to what is stated in the documentation, the confidence interval
     size `conf_sig` was 0.6 by default instead of 0.9.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
521
522
523
524


 - Conditional forecasts

525
526
527
   + The `conditional_forecast` command produced wrong results in calibrated
     models when used at initial values outside of the steady state (given with
     `initval`),
Stéphane Adjemian's avatar
Stéphane Adjemian committed
528

529
530
   + The `plot_conditional_forecast` option could produce unreadable figures if
     the areas overlap,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
531

532
   + The `conditional_forecast` command after MLE crashed,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
533

534
535
   + In contrast to what is stated in the manual, the confidence interval size
     `conf_sig` was 0.6 by default instead of 0.8.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
536

537
538
539
   + Conditional forecasts were wrong when the declaration of endogenous
     variables was not preceeding the declaration of the exogenous
     variables and parameters.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
540
541
542
543


 - Discretionary policy

544
545
   + Dynare allowed running models where the number of instruments did not match
     the number of omitted equations,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
546

547
   + Dynare could crash in some cases when trying to display the solution,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
548

549
550
   + Parameter dependence embedded via a `steady_state` was not taken into
     account, typically resulting in crashes.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
551
552
553

 - dseries class

554
555
   + When subtracting a dseries object from a number, the number was instead
     subtracted from the dseries object.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
556
557
558
559


 - DSGE-VAR models

560
561
   + Dynare crashed when estimation encountered non-finite values in the Jacobian
     at the steady state,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
562

563
564
565
566
   + 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.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
567
568
569
570


 - Estimation command

571
572
   + 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,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
573

574
   + Calling estimation after identification could lead to crashes,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
575

576
577
578
579
   + 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`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
580

581
582
   + Computation of Bayesian smoother could crash for larger models when
     requesting `forecast` or `filtered_variables`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
583

584
585
   + Geweke convergence diagnostics were not computed on the full MCMC chain when
     the `load_mh_file` option was used,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
586

587
588
   + The Geweke convergence diagnostics always used the default `taper_steps` and
   `geweke_interval`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
589

590
591
   + Bayesian IRFs (`bayesian_irfs` option) could be displayed in an unreadable
     way when they move from negative to positive values,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
592

593
594
   + If `bayesian_irfs` was requested when `mh_replic` was too low to compute
     HPDIs, plotting was crashing,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
595

596
597
598
   + 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_*`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
599

600
   + Using a user-defined `mode_compute` crashed estimation,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
601

602
   + Option `mode_compute=10` did not work with infinite prior bounds,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
603

604
605
606
   + 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,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
607

608
609
   + Using the `forecast` option with `loglinear` erroneously added the unlogged
     steady state,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
610

611
612
   + When using the `loglinear` option the check for the presence of a constant
     was erroneously based on the unlogged steady state,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
613

614
615
   + Estimation of `observation_trends` was broken as the trends specified as a
     function of deep parameters were not correctly updated during estimation,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
616

617
618
619
   + When using `analytic_derivation`, the parameter values were not set before
     testing whether the steady state file changes parameter values, leading to
     subsequent crashes,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
620

621
622
623
   + 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,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
624

625
626
   + When computing posterior moments, Dynare falsely displayed that moment
     computations are skipped, although the computation was performed correctly,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
627

628
629
630
   + If `conditional_variance_decomposition` was requested, although all
     variables contain unit roots, Dynare crashed instead of providing an error
     message,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
631

632
633
634
   + Computation of the posterior parameter distribution was erroneously based
     on more draws than specified (there was one additional draw for every Markov
     chain),
Stéphane Adjemian's avatar
Stéphane Adjemian committed
635

636
   + The estimation option `lyapunov=fixed_point` was broken,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
637

638
   + Computation of `filtered_vars` with only one requested step crashed Dynare,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
639

640
   + Option `kalman_algo=3` was broken with non-diagonal measurement error,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
641

642
643
   + When using the diffuse Kalman filter with missing observations, an additive
     factor log(2*pi) was missing in the last iteration step,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
644

645
646
   + Passing of the `MaxFunEvals` and `InitialSimplexSize` options to
     `mode_compute=8` was broken,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
647

648
649
   + Bayesian forecasts contained initial conditions and had the wrong length in
     both plots and stored variables,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
650

651
652
653
   + 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,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
654

655
   + Computation of smoothed measurement errors in Bayesian estimation was broken,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
656

657
658
659
   + The `selected_variables_only` option (`mh_replic=0`, ML, or
     `calibrated_smoother`) returned wrong results for smoothed, updated, and
     filtered variables,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
660

661
662
   + Combining the `selected_variables_only` option with forecasts obtained
     using `mh_replic=0`, ML, or `calibrated_smoother` leaded to crashes,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
663

664
   + `oo_.UpdatedVariables` was only filled when the `filtered_vars` option was specified,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
665

666
667
668
   + 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`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
669

670
671
   + 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,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
672

673
674
675
676
677
678
   + 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`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
679

680
681
682
   + 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,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
683

684
685
   + When the initialization of an MCMC run failed, the metropolis.log file was
     locked, requiring a restart of Matlab to restart estimation,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
686

687
688
   + If the posterior mode was right at the corner of the prior bounds, the
     initialization of the MCMC erroneously crashed,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
689

690
691
692
   + 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.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
693
694
695
696


 - Estimation and calibrated smoother

697
698
   + When using `observation_trends` with the `prefilter` option, the mean shift
     due to the trend was not accounted for,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
699

700
701
702
   + 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,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
703

704
705
   + The diffuse Kalman smoother was crashing if the forecast error variance
     matrix becomes singular,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
706

707
708
   + The multivariate Kalman smoother provided incorrect state estimates when
     all data for one observation are missing,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
709

710
711
   + The multivariate diffuse Kalman smoother provided incorrect state estimates
     when the `Finf` matrix becomes singular,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
712

713
714
   + The univariate diffuse Kalman filter was crashing if the initial covariance
     matrix of the nonstationary state vector is singular,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
715
716
717
718


 - Forecats

719
720
   + In contrast to what is stated in the manual, the confidence interval size
     `conf_sig` was 0.6 by default instead of 0.9.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
721

722
723
   + Forecasting with exogenous deterministic variables provided wrong decision
     rules, yielding wrong forecasts.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
724

725
726
   + Forecasting with exogenous deterministic variables crashed when the
     `periods` option was not explicitly specified,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
727

728
729
730
   + 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.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
731
732
733
734


 - Global Sensitivity Analysis

735
   + Sensitivity with ML estimation could result in crashes,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
736

737
   + Option `mc` must be forced if `neighborhood_width` is used,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
738

739
   + Fixed dimension of `stock_logpo` and `stock_ys`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
740

741
   + Incomplete variable initialization could lead to crashes with `prior_range=1`.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
742
743
744
745


 - Indentification

746
747
748
   + Identification did not correctly pass the `lik_init` option,
     requiring the manual setting of `options_.diffuse_filter=1` in
     case of unit roots,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
749

750
751
   + Testing identification of standard deviations as the only
     parameters to be estimated with ML leaded to crashes,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
752

753
754
755
   + Automatic increase of the lag number for autocovariances when the
     number of parameters is bigger than the number of non-zero moments
     was broken,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
756

757
   + When using ML, the asymptotic Hessian was not computed,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
758

759
760
   + Checking for singular values when the eigenvectors contained only
     one column did not work correctly,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
761
762
763
764


 - Model comparison

765
   + Selection of the `modifiedharmonicmean` estimator was broken,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
766
767
768
769


 - Optimal Simple Rules

770
771
772
   + When covariances were specified, variables that only entered with
     their variance and no covariance term obtained a wrong weight,
     resulting in wrong results,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
773

774
775
776
777
   + 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,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
778

779
   + Using only one (co)variance in the objective function resulted in crashes,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
780

781
   + For models with non-stationary variables the objective function was computed wrongly.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
782
783
784
785


 - Ramsey policy

786
787
   + 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.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
788

789
790
   + When using an external steady state file, incorrect steady states
     could be accepted,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
791

792
793
   + When using an external steady state file with more than one
     instrument, Dynare crashed,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
794

795
796
   + When using an external steady state file and running `stoch_simul`
     after `ramsey_planner`, an incorrect steady state was used,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
797

798
799
   + When the number of instruments was not equal to the number of
     omitted equations, Dynare crashed with a cryptic message,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
800

801
   + The `planner_objective` accepted `varexo`, but ignored them for computations,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
802
803
804
805


 - Shock decomposition

806
807
   + Did not work with the `parameter_set=calibration` option if an
     `estimated_params` block is present,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
808

809
   + Crashed after MLE.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
810
811
812
813


 - Perfect foresight models

814
815
   + The perfect foresight solver could accept a complex solution
     instead of continuing to look for a real-valued one,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
816

817
   + The `initval_file` command only accepted column and not row vectors,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
818

819
   + The `initval_file` command did not work with Excel files,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
820

821
822
823
   + Deterministic simulations with one boundary condition crashed in
     `solve_one_boundary` due to a missing underscore when passing
     `options_.simul.maxit`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
824

825
826
   + Deterministic simulation with exogenous variables lagged by more
     than one period crashed,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
827

828
829
   + Termination criterion `maxit` was hard-coded for `solve_algo=0`
     and could no be changed,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
830

831
   + When using `block`/`bytecode`, relational operators could not be enforced,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
832

833
834
   + When using `block` some exceptions were not properly handled,
     leading to code crashes,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
835

836
   + Using `periods=1` crashed the solver (bug only partially fixed).
Stéphane Adjemian's avatar
Stéphane Adjemian committed
837
838
839
840


 - Smoothing

841
842
   + The univariate Kalman smoother returned wrong results when used
     with correlated measurement error,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
843

844
845
846
   + The diffuse smoother sometimes returned linear combinations of the
     smoothed stochastic trend estimates instead of the original trend
     estimates.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
847
848
849

 - Perturbation reduced form

850
851
852
   + 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`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
853

854
855
   + Dynare could crash when the steady state could not be computed
     when using the `loglinear` option,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
856

857
858
   + Using `bytcode` when declared exogenous variables were not
     used in the model leaded to crashes in stochastic simulations,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
859

860
861
   + Displaying decision rules involving lags of auxiliary variables of
     type 0 (leads>1) crashed.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
862

863
864
   + The `relative_irf` option resulted in wrong output at `order>1` as
     it implicitly relies on linearity.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914


 - 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.


915

916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
Announcement for Dynare 4.4.3 (on 2014-07-31)
=============================================

We are pleased to announce the release of Dynare 4.4.3.

This is a bugfix release.

The Windows packages are already available for download at:

 http://www.dynare.org/download/dynare-stable

The Mac and GNU/Linux packages (for Debian and Ubuntu LTS) should follow soon.

This release is compatible with MATLAB versions 7.3 (R2006b) to 8.2 (R2013b)
and with GNU Octave versions 3.6 to 3.8.

Here is a list of the problems identified in version 4.4.2 and that have been
fixed in version 4.4.3:

 - When loading a dataset in XLS, XLSX or CSV format, the first
   observation was discarded.

 - Reading data in an Excel-file with only one variable wasz leading
   to a crash.

 - When using the k_order_perturbation option (which is implicit at
   3rd order) without the use_dll option, crashes or unexpected
   behavior could happen if some 2nd or 3rd derivative evaluates to
   zero (while not being symbolically zero)

 - When using external function, Ramsey policy could crash or return
   wrong results.

 - For Ramsey policy, the equation numbers associated with the
   Lagrange multipliers stored in M_.aux_vars were erroneously one too
   low

 - When updating deep parameters in the steady state file, the changes
   were not fully taken into account (this was only affecting the
   Ramsey policy).

 - When using external functions and the bytecode option, wrong
   results were returned (if second order derivates of the external
   functions were needed).

 - The confidence level for computations in estimation, conf_sig could
   not be changed and was fixed at 0.9. The new option mh_conf_sig is
   now used to set this interval

 - Conditional forecasts with non-diagonal covariance matrix used an
   incorrect decomposition of the covariance matrix. A Cholesky
   factorization is used.

 - Option geweke_interval was not effective, Dynare always defaulted
   to the standard value.

 - The mode_file option lacked backward compatibility with older
   Dynare versions.

 - Loading an mh_mode file with the mode_file option was broken.

 - Using identification with var_exo_det leaded to crashes (the
   preprocessor now returns an error if they are used simultaneously)

 - The identification command did not print results if the initial
   parameter set was invalid and then crashed later on if the MC
   sample is bigger than 1

 - Inconsistencies between static and dynamic models leaded to crashes
   instead of error messages (only with block option).

 - The use of external functions crashed the preprocessor when the
   derivatives of the external function are explicitly called in the
   model block. The preprocessor now forbids the use of external
   functions derivates in the model block.

 - Using the block option when a variable does not appear in the
   current period crashed Dynare instead of providing an error
   message.


997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
Announcement for Dynare 4.4.2 (on 2014-03-04)
=============================================

We are pleased to announce the release of Dynare 4.4.2.

This is a bugfix release.

The Windows packages are already available for download at:

 http://www.dynare.org/download/dynare-stable

The Mac and GNU/Linux packages (for Debian and Ubuntu LTS) should follow soon.

This release is compatible with MATLAB versions 7.3 (R2006b) to 8.2 (R2013b)
and with GNU Octave versions 3.6 to 3.8.

Here is a list of the problems identified in version 4.4.1 and that have been
fixed in version 4.4.2:

 - Geweke convergence diagnostics was computed on the wrong sample if `mh_drop'
   was not equal to the default of 0.5.

 - The `loglinear' option of `stoch_simul' was displaying the steady state of
   the original values, not the logged ones, and was producing incorrect
   simulations and simulated moments. Theoretical moments were unaffected.

 - The `optim' option of `estimation (for setting options to `mode_compute')
   was only working with at least MATLAB 8.1 (R2013a) or Octave 3.8.

 - For unit root models, theoretical HP filtered moments were sometimes
   erroneously displayed as NaN.

 - Specifying an endogenous variable twice after the `estimation' command would
   lead to a crash in the computation of moments.

 - Deterministic simulations were crashing on some models with more than one
   lead or one lag on exogenous variables.

 - Homotopy in stochastic extended path with order greater than 0 was not
   working correctly (during the homotopy steps the perfect foresight model
   solver was called instead of the stochastic perfect foresight model solver).

 - MCMC convergence diagnostics were not computed if `mh_replic' was less than
   2000; the test now relies on the total number of iterations (this only makes
   a difference if option `load_mh_file' is used).


1044
Announcement for Dynare 4.4.1 (on 2014-01-17)
Houtan Bastani's avatar
Houtan Bastani committed
1045
1046
1047
1048
=============================================

We are pleased to announce the release of Dynare 4.4.1.

1049
1050
This release contains a few changes to the user interface and fixes various
bugs. It also adds compatibility with Octave 3.8.
Houtan Bastani's avatar
Houtan Bastani committed
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060

The Windows packages are already available for download at:

 http://www.dynare.org/download/dynare-stable

The Mac and GNU/Linux packages (for Debian and Ubuntu) should follow soon.

All users are encouraged to upgrade.

This release is compatible with MATLAB versions 7.3 (R2006b) to 8.2 (R2013b) and
Sébastien Villemot's avatar
Sébastien Villemot committed
1061
with GNU Octave versions 3.6 to 3.8.
Houtan Bastani's avatar
Houtan Bastani committed
1062

1063
* Changes to the user interface:
Houtan Bastani's avatar
Houtan Bastani committed
1064
1065

 - The syntax introduced in 4.4.0 for conditional forecast in a deterministic
1066
   setup was removed, and replaced by a new one that is better suited to the
1067
   task. More precisely, such deterministic forecasts are no longer done using
1068
1069
1070
   the `conditional_forecast' command. The latter is replaced by a group of
   commands: `init_plan', `basic_plan' and `flip_plan'. See the reference
   manual for more details.
Houtan Bastani's avatar
Houtan Bastani committed
1071

1072
1073
1074
1075
1076
1077
1078
1079
 - Changes to the reporting module: option `annualAverages' to `addTable' has
   been removed (use option `tableDataRhs' to `addSeries' instead); option
   `vlineAfter' to `addTable' now also accepts a cell array.

 - Changes to the date and time series classes: implement broadcasting for
   operations (+,-,* and /) between `dseries' class and scalar or vectors; add
   the possibility of selecting an observation within a time series using a
   formatted string containing a date.
Houtan Bastani's avatar
Houtan Bastani committed
1080
1081
1082
1083
1084
1085
1086

* Bugs and problems identified in version 4.4.0 and that have been fixed in
  version 4.4.1:

 - In MS-SBVAR, there was a bug preventing the computation of impulse responses
   on a constant regime.

1087
1088
 - Under Octave, after modifying the MOD file, the changes were not taken into
   account at the first Dynare run, but only at the second run.
1089

Houtan Bastani's avatar
Houtan Bastani committed
1090

1091
Announcement for Dynare 4.4.0 (on 2013-12-16)
1092
1093
1094
1095
1096
1097
=============================================

We are pleased to announce the release of Dynare 4.4.0.

This major release adds new features and fixes various bugs.

1098
1099
1100
1101
1102
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.
1103
1104
1105
1106

All users are strongly encouraged to upgrade.

This release is compatible with MATLAB versions ranging from 7.3 (R2006b) to
1107
8.2 (R2013b) and with GNU Octave version 3.6.
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184

Here is the list of major user-visible changes:


* New major algorithms:

 - Extended path at order 1 and above, also known as “stochastic extended
   path”. This method is triggered by setting the `order' option of the
   `extended_path' command to a value greater than 0. Dynare will then use a
   Gaussian quadrature to take into account the effects of future uncertainty.
   The time series for the endogenous variables are generated by assuming that
   the agents believe that there will no more shocks after period t+order.

 - Alternative algorithms for computing decision rules of a stochastic model,
   based on the cycle reduction and logarithmic reduction algorithms. These
   methods are respectively triggered by giving `dr = cycle_reduction' or 'dr
   = logarithmic_reduction' as an option to the `stoch_simul' command.

 - Pruning now works with 3rd order approximation, along the lines of
   Andreasen, Fernández-Villaverde and Rubio-Ramírez (2013).

 - Computation of conditional forecast using an extended path method. This is
   triggered by the new option `simulation_type = deterministic' in the
   `conditional_forecast' command. In this case, the `expectation' command in
   the `conditional_forecast_paths' block has to be used to indicate the nature
   of expectations (whether shocks are a surprise or are perfectly
   anticipated).

 - Endogenous priors as in Christiano, Trabandt and Walentin (2011). Those are
   triggered by the new option `endogenous_prior' of the `estimation' command.


* Other algorithmic improvements:

 - New command `model_diagnostics' to perform various sanity checks on the
   model. Note: in the past, some users may have used a preliminary MATLAB
   function implementing this; the new command has the same syntax, except that
   you shouldn't pass any argument to it.

 - Terminal conditions of perfect foresight simulations can now be specified in
   growth rates. More specifically, the new option `differentiate_forward_vars'
   of the `model' block will create auxiliary forward looking variables
   expressed in first differences or growth rates of the actual forward looking
   variables defined in the model. These new variables have obvious zero
   terminal conditions whatever the simulation context and this in many cases
   helps convergence of simulations.

 - Convergence diagnostics for single chain MCMC à la Geweke (1992, 1999).

 - New optimizer for the posterior mode (triggered by `mode_compute=10'): it