NEWS.md 106 KB
Newer Older
1
Announcement for Dynare 4.6.3 (on 2020-11-23)
Sébastien Villemot's avatar
Sébastien Villemot committed
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
=============================================

We are pleased to announce the release of Dynare 4.6.3.

This maintenance release fixes various bugs.

The Windows, macOS and source packages are already available for download at
[the Dynare website](https://www.dynare.org/download/).

All users are strongly encouraged to upgrade.

This release is compatible with MATLAB versions ranging from 7.9 (R2009b) to
9.9 (R2020b), and with GNU Octave versions 5.2.0 (under Windows) and 4.4.1
(under macOS).

Here is a list of the problems identified in version 4.6.2 and that have been
fixed in version 4.6.3:

* Using an unknown symbol in `irf_shocks` option of `stoch_simul` would lead to
  a crash of the preprocessor
22
* `identification` would crash for purely forward-looking models
Sébastien Villemot's avatar
Sébastien Villemot committed
23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
* The `endogenous_prior` option did not properly handle missing observations
* The auxiliary particle filter with pruning and resampling would crash
* Initialization of the state variance for particle filters was buggy
* An `@#else` clause after an `@#ifndef` was not correctly interpreted
* An `@#elseif` clause after an `@#ifdef` or an `@#ifndef` was not correctly
  interpreted
* Perfect foresight simulations of models with a single equation would crash
  when using either the `lmmcp` option or the `linear_approximation`
* Inequality constraints on endogenous variables (when using the `lmmcp`
  option) were not enforced on purely backward or purely forward models
* Perfect foresight simulations with `bytecode` and `block` options could crash
  if there was a purely forward variable whose value in all periods could be
  evaluated backward (typically a process of the form `y=a*y(+1)+e`)
* `extended_path` was broken with `bytecode`
* Under Windows, with Octave, the k-order perturbation and MS-SBVAR MEX files
  could not be loaded
* On Fedora (and possibly other GNU/Linux distributions), compilation from
  source would fail against Octave 5

Sébastien Villemot's avatar
Sébastien Villemot committed
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 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
Announcement for Dynare 4.6.2 (on 2020-09-07)
=============================================

We are pleased to announce the release of Dynare 4.6.2.

This maintenance release fixes various bugs.

The Windows, macOS and source packages are already available for download at
[the Dynare website](https://www.dynare.org/download/).

All users are strongly encouraged to upgrade.

This release is compatible with MATLAB versions ranging from 7.9 (R2009b) to
9.8 (R2020a), and with GNU Octave versions 5.2.0 (under Windows) and 4.4.1
(under macOS).

*Note for Windows users:* upon launching the Dynare installer, you may get a
warning emitted by Windows Defender SmartScreen, saying that this is an
unrecognized app and that it was prevented from starting. You can safely ignore
this warning, as long as you can verify on the next screen that CEPREMAP is the
editor of the software. This security warning is due to the fact that we had to
renew our code signing certificate (which had expired), and it takes some time
to rebuild our reputation as a software editor using the new certificate.

Here is a list of the problems identified in version 4.6.1 and that have been
fixed in version 4.6.2:

* Perfect foresight simulations of purely backward models could deliver an
  incorrect result if some exogenous variable appeared with a lag of 2 or more
  (and neither `block` nor `bytecode` option was used)
* Perfect foresight simulations of linear models could deliver an incorrect
  result if the following four conditions were met:
  + the model was actually declared as linear through the `linear` option
  + there was an exogenous variable with a lead or a lag
  + `stack_solve_algo` was equal to 0 (the default) or 7
  + neither `block` nor `bytecode` option was used
* In stochastic simulations, for variables that actually do not leave the
  steady state, reported simulated moments could be spurious (due to division
  by zero)
* Displayed variance decompositions would only take into account measurement
  errors if measurement errors were present for all observed variables
* The posterior variance decompositions with measurement errors computed with
  `moments_varendo` were incorrect
* `moments_varendo` would not update `oo_.PosteriorTheoreticalMoments` if it
  was already present, from *e.g.* an earlier run of `estimation`
* Identification would in some cases compute wrong Jacobian of moments
* Identification would display incorrect results if parameter dependence was
  implemented via a steady state file
* `generate_trace_plots` would crash when measurement errors were present
* `estimation` would crash for correlated measurement errors
* Parallel execution/testing could crash instead of aborting with a proper
  error message
* Under macOS, Dynare would incorrectly claim that it is compiled for Octave
  5.2.0 (it is actually compiled for Octave 4.4.1)
* Using external functions in a model local variable would crash the
  preprocessor
* Tolerance criteria for steady state computations were inconsistently set
* `stoch_simul` with its default `order=2` would crash with a message about
  `hessian_eq_zero` not existing if an explicit `order=1` was present somewhere
  else in the `.mod` file
* Model local variables were not written to the `modfile.json` JSON file
* Model local variables names would have two spurious underscores at their
  point of definition in the `dynamic.json` and `static.json` files (but only
  in the definition, not when they were used, which is inconsistent)
* The `solve_algo=9` option was not accessible. The `solve_algo=10` and
  `solve_algo=11` options were not accessible with `block` (without `bytecode`)
* Under certain circumstances, `extended_path` would crash when used in
  conjunction with the `block` option
* `extended_path` was not working with the `bytecode` option
* `shock_decomposition` was not accepting the options of `estimation` related
  to smoothing
* `conditional_forecast` would display a warning even if the simulation was
  successful
* The `prior_trunc` option of `identification` was not working
* The `rand_multivariate_student` value of the `proposal_distribution` option
  was not working when used with the
  `tailored_random_block_metropolis_hastings` posterior sampling method
* Perfect foresight simulations of backward models would crash if convergence
  failed with complex-valued residuals
* The diffuse Kalman smoother would crash if `Finf` became singular

Sébastien Villemot's avatar
Sébastien Villemot committed
123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165
Announcement for Dynare 4.6.1 (on 2020-03-13)
=============================================

We are pleased to announce the release of Dynare 4.6.1.

This maintenance release fixes various bugs.

The Windows, macOS and source packages are already available for download at
[the Dynare website](https://www.dynare.org/download/).

All users are strongly encouraged to upgrade.

This release is compatible with MATLAB versions ranging from 7.9 (R2009b) to
9.7 (R2019b), and with GNU Octave versions 5.2.0 (under Windows) and 4.4.1
(under macOS).

Here is a list of the problems identified in version 4.6.0 and that have been
fixed in version 4.6.1:

* Installation on macOS would fail if the GCC compiler was supposed to be
  installed and `www.google.com` was not reachable or blocked
* Dynare++ was missing the `dynare_simul.m` file
* The parameter vector `M_.params` would not be correctly updated after calls
  to `stoch_simul` and `discretionary_policy` if parameters had been modified
  in a steady state file
* The `stoch_simul` command with both the `nograph` and `TeX` options would
  crash
* The `stoch_simul` command with the `noprint` option would crash
* The `prior moments` command would crash if the used parameter vector
  triggered an error code
* In case of problem, the `discretionary_policy` command would crash instead of
  aborting with a proper error code
* Computing of prior/posterior statistics would not work in parallel
* Closing of parallel estimation on GNU/Linux could crash
* The `histval` command would not work in combination with the
  `predetermined_variables` command
* Ramsey optimal policy with multiple instruments would crash if a steady state
  file returned complex values, instead of providing an error message
* The `model_diagnostics` command would not correctly update the parameter
  vector if the latter was set in a steady state file
* The `model_diagnostics` command would ignore the `nocheck` steady state flag


166 167
Announcement for Dynare 4.6.0 (on 2020-02-20)
=============================================
168 169 170 171 172 173 174 175 176 177 178 179 180 181

We are pleased to announce the release of Dynare 4.6.0.

This major release adds new features and fixes various bugs.

The Windows, macOS and source packages are already available for download at
[the Dynare website](https://www.dynare.org/download/).

All users are strongly encouraged to upgrade.

This release is compatible with MATLAB versions ranging from 7.9 (R2009b) to
9.7 (R2019b), and with GNU Octave versions 5.2.0 (under Windows) and 4.4.1
(under macOS).

182 183
Major user-visible changes
--------------------------
184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330

 - Stochastic simulations

    - The perturbation method is now available at an arbitrary approximation
      order. In other words, the `order` option of `stoch_simul` accepts an
      arbitrary positive integer (of course, up to some model-specific
      computational limit).

    - New option `filtered_theoretical_moments_grid` of `stoch_simul`, that
      supersedes `hp_ngrid`.

 - Estimation

    - Nonlinear estimation is now also available at an arbitrary approximation
      order. In other words, the `order` option of `estimation` accepts an
      arbitrary positive integer (of course, up to some model-specific
      computational limit).

    - Various improvements to particle filters.

    - It is now possible to estimate models under optimal policy (see below).

    - Variance decomposition of observables now accounts for measurement error.

    - New option `mh_tune_jscale` of `estimation` command for tuning the scale
      parameter of the proposal distribution of the Random Walk Metropolis
      Hastings.

    - Added debugging info when parameters take a `NaN` or `Inf` value.

    - Option `mode_compute=1` is now available under Octave.

 - Perfect foresight and extended path

    - A significant speed improvement should be noted on large models (when
      neither `bytecode` nor `block` option is used). The stacked problem is
      now constructed using a dedicated machine-compiled library that greatly
      speeds up the process (in particular, the time spent in that step can
      become negligible when the `use_dll` option is used).

    - New options `print` and `noprint` of `perfect_foresight_solver` command.

    - Option `stack_solve_algo=2` is now available under Octave.

 - Steady state

    - Option `solve_algo=7` is now available under Octave.

 - Optimal policy

    - The `ramsey_policy` command is now deprecated. It is superseded by
      successive calls to `ramsey_model`, `stoch_simul`, and
      `evaluate_planner_objective` (in this order).

    - It is now possible to estimate a model under optimal policy (either
      Ramsey or discretionary) by running the `estimation` command after either
      `ramsey_model` or `discretionary_policy`. It is however not yet possible
      to estimate parameters that appear in the discount factor of the social
      planner.

    - Discretionary policy returns a more informative error message when the
      objective has nonzero derivatives with respect to some variables.

 - Identification

    - Added minimal system identification check of *Komunjer and Ng (2011)*.

    - Added spectrum identification check of *Qu and Tkachenko (2012)*.

    - Identification is now also available for approximation orders 2 and 3
      with either analytical or numerical parameter derivatives. The relevant
      moments and spectrum are computed from the pruned state space system
      as in *Mutschler (2015)*.

    - All tests (moments, spectrum, minimal system, strength) can be turned
      off.

    - More numerical options can be changed by the user.

    - Improved printing and storage (same folder) of results.

 - Sensitivity analysis

    - New `diffuse_filter` option to the `dynare_sensitivity` command.

    - Arbitrary expressions can now be passed for the interval boundaries in
      `irf_calibration` and `moment_calibration`. ⚠ This breaks the
      previous syntax, requiring that the lower/upper bounds be separated by
      commas.

 - Forecasting and smoothing

    - In `conditional_forecast_paths`, it is no longer required that all
      constrained paths be of the same length. There may now be a different
      number of controlled variables at each period. In that case, the order of
      declaration of endogenous controlled variables and of `controlled_varexo`
      matters: if the second endogenous variable is controlled for less periods
      than the first one, the second `controlled_varexo` isn't set for the last
      periods.

    - New option `parameter_set` to the `calib_smoother` command.

    - ⚠ The results of `conditional_forecast` command is now saved in
      `oo_` (used to be in a file)

 - Shock decomposition

   - Added `fast_realtime` option to real time shock decomposition (deactivated
     by default, runs the smoother only twice: once for the last in-sample and
     once for the last out-of-sample data point).

   - New `diff`, `flip`, `max_nrows`, `plot_init_date` and `plot_end_date`
     options to `plot_shock_decomposition`.

   - New `initial_decomposition_decomposition` command, for computing and
     plotting the decomposition of the effect of smoothed initial conditions of
     state variables.

   - New `squeeze_shock_decomposition` command, for removing decompositions of
     variables that are not of interest.

   - New `with_epilogue` option (common to `shock_decomposition`,
     `realtime_shock_decomposition` and `initial_condition_decomposition`).

   - New `init2shocks` block to attribute initial conditions to shocks.

 - Macro processor

   - New object types: real (supersedes integers), boolean (distinct from
     integers), tuple, user-defined function.

   - New operators: various mathematical functions, set operations on arrays
     (union, intersection, difference, cartesian power and product), type
     checking and conversion.

   - Added support for comprehensions (*e.g.* the set containing the squares of
     all even numbers between 1 and 5 can be constructed with `[ i^2 for i in
     1:5 when mod(i,2) == 0]`).

   - User-defined functions can be declared using the `@#define` operator (*e.g.*
     `@#define f(x) = 2*x^2+3*x+5`).

   - `@#elseif`-clauses are now supported in conditional statements.

   - `@#for` loops can iterate over several variables at the same time (*e.g.*
     `@#for (i,j) in X`, where `X` is an array containing tuples of size 2).

331 332 333 334
   - Added the possibility to exclude some elements when iterating over `@#for`
     loops (*e.g.* `@#for i in 1:5 when mod(i,2) == 0` iterates over all even
     numbers between 1 and 5).

335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494
   - A `defined()` function allows testing whether macro variables have been
     defined.

   - Empty arrays (with the `[]` syntax) are now possible.

   - Arrays of arrays are now supported.

   - New macro directives `@#echomacrovars` and `@#echomacrovars(save)` for
     displaying or saving the values of all macro-variables.

   - Inline comments are now supported.

   - ⚠ All division operations are now done with doubles (as opposed to
     integers). To achieve the old functionality, use the new `floor` operator.

   - ⚠ Colon syntax used to require braces around it to create an array
     (*e.g.* `[1:3]` would create `[1,2,3]`). Now this is not necessary (`1:3`
     creates `[1,2,3]` while `[1:3]` would create `[[1,2,3]]`).

   - ⚠ Previously, printing a boolean would print `1` or `0`. Now, it
     prints `true` or `false`. To achieve the old functionality, you must cast
     it to a real, *e.g.* `@{(real)(1!=0)}`.

 - LaTeX output

    - New command `write_latex_steady_state_model`.

    - New option `planner_discount_latex_name` of `ramsey_model` and
      `discretionary_policy`.

    - New command `model_local_variable` command for assigning a LaTeX name to
      model-local variables.

    - The `write_latex_static_model` and `write_latex_original_model` commands
      now support the `write_equation_tags` option.

 - Compilation of the model (`use_dll` option) made easier and faster

    - Under Windows, it is no longer necessary to manually install the
      compiler, since the latter is now shipped by the Dynare installer.

    - Under macOS, the Dynare installer now automatically downloads and
      installs the compiler.

    - It is no longer necessary to configure MATLAB to let it know where the
      compiler is, since the compilation is now done by the preprocessor.

    - The compilation phase is now faster on large models (this has been
      achieved by disabling a few time-consuming and not-so-useful optimization
      passes otherwise done by the compiler).

    - New `compilation_setup` block for specifying a custom compiler or custom
      compilation flags.

 - Model, variables and parameters declaration

    - New syntax to declare model variables and parameters on-the-fly in the
      `model` block. To do this, simply follow the symbol name with a vertical
      line (`|`, pipe character) and either an `e`, an `x`, or a `p`. For
      example, to declare a parameter named `alpha` in the model block, you
      could write `alpha|p` directly in an equation where it appears.
      Similarly, to declare an endogenous variable `c` in the model block you
      could write `c|e`.

    - New syntax to declare model variable and parameters on-the-fly in
      equation tags. In the tag, simply state the type of variable to be
      declared (`endogenous`, `exogenous`, or `parameter` followed by an equal
      sign and the variable name in single quotes. Hence, to declare a variable
      `c` as endogenous in an equation tag, you can type `[endogenous='c']`.

    - New `epilogue` block for computing output variables of interest that may
      not be necessarily defined in the model (*e.g.* various kinds of
      real/nominal shares or relative prices, or annualized variables out of a
      quarterly model).

 - Command-line options

    - Added the possibility to declare Dynare command-line options in the `.mod`
      file.

    - New option `nopreprocessoroutput` to disable printing of messages from
      the preprocessor.

    - It is now possible to assign an arbitrary macro-expression to a
      macro-variable defined on the command-line, using the `-D` syntax.

    - New  option `linemacro` to revert to the old format of the
      macro-processed file (see below).

 - Preprocessor outputs and inputs

    - Added JSON output to the preprocessor. A representation of the model file
      and the whole content of the `.mod` file is saved in `.json` files.
      These JSON files can be easily parsed from any language (C++, Fortran,
      Python, Julia, MATLAB, Octave…). This new feature opens the possibility to
      develop alternative back-ends for the Dynare language.

    - ⚠ Most files generated by the preprocessor are now grouped under
      two subdirectories. Assuming your file is `FILENAME.mod`, then M-files
      and MEX-files will be under `+FILENAME/`, while other output (JSON,
      LaTeX, source code for the MEX files) will be under `FILENAME/`.

    - The macro-generated output is now more readable (no more line numbers and
      empty lines). The old behaviour can be restored using the `linemacro`
      option (see above).

    - Ability to call the preprocessor by passing the `.mod` file as a string
      argument from the macOS or GNU/Linux command line.

 - dseries classes

    - New functionalities and efficiency improvements.

    - Complete rewrite using the new `classdef` syntax and exploiting in place
      modifications when possible.

    - Integration of the `dates` classes within `dseries`.

 - Reporting classes

    - Automatically create titlepage with page numbers/page titles.

    - Allow for the removal of headers and footers from a given page.

    - Allow user to set page number.

    - Split up report output. Create new files for the preamble, the body of
      the report, and each individual page of the report.

    - The classes have been converted to the new `classdef` syntax.

 - Misc

    - External functions can be located in MATLAB/Octave namespaces.

    - Improvements to the balanced growth path test that is performed after
      Dynare has detrended the model (given the trends on variables declared by
      the user): the default tolerance has been raised, and a different value
      can be set with new option `balanced_growth_test_tol` to the `model`
      block; as a consequence, failing the test is now an error again.

    - New collection of MATLAB/Octave utilities to retrieve and alter objects:
      `get_irf`, `get_mean`, `get_shock_stderr_by_name`, `get_smooth`,
      `get_update`, `set_shock_stderr_value`.

    - ⚠ Previously, when some MEX files were missing, Dynare would
      automatically fall back to slower M-file functional alternative; this is
      no longer the case. It is however still possible to manually add these
      alternatives in the MATLAB/Octave path (they are located under
      `matlab/missing/mex`; this only applies to the `mjdgges`, `gensylv`,
      `A_times_B_kronecker_C`, `sparse_hessian_times_B_kronecker_C` and
      `local_state_space_iteration_2` DLLs).


Since there are a few backward-incompatible changes in this release, users may
want to have a look at the [upgrade
guide](https://git.dynare.org/Dynare/dynare/-/wikis/BreakingFeaturesIn4.6) to
adapt their existing codes.


495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540
Bugs that were present in 4.5.7 and that are fixed in 4.6.0
-----------------------------------------------------------

* Estimation: the check for stochastic singularity erroneously would only take
  estimated measurement error into account.
* Estimation: if the Hessian at the mode was not positive definite, the Laplace
  approximation returned a complex number, but only displayed the real-valued
  part.
* Conditional Forecasting: using one period only would result in a crash.
* First-order approximation was not working with purely forward-looking models.
* The preprocessor would not allow for inline comments including macro
  statements.
* Using the `STEADY_STATE()` operator on exogenous variables would lead to
  crashes in stochastic simulations.
* `moment_calibration`: for autocorrelation functions, the x-axis labeling had
  the wrong order.
* `plot_identification`: placement of white dots indicating infinite values was
  incorrect
* Automatic detrending would sometime refuse to detrend model despite the user
  having given correct trends.
* Using `use_dll` + `fast` options would not always recompile the model when
  the equations were changed.
* Under certain circumstances, the combination of `bytecode` and
  `stack_solve_algo=1` options could lead to crashes or wrong results.


References
----------

 - Komunjer, I. and S. Ng (2011), “[Dynamic Identification of Dynamic
   Stochastic General Equilibrium
   Models](https://www.onlinelibrary.wiley.com/doi/abs/10.3982/ECTA8916),”
   *Econometrica*, 79(6), 1995–2032

 - Qu, Z. and D. Tkachenko (2012), “[Identification and frequency domain
   quasi‐maximum likelihood estimation of linearized dynamic stochastic
   general equilibrium
   models](https://onlinelibrary.wiley.com/doi/abs/10.3982/QE126),”
   *Quantitative Economics*, 3(1), 95–132

 - Mutschler, W. (2015), “[Identification of DSGE models—The effect of
   higher-order approximation and
   pruning](https://www.sciencedirect.com/science/article/pii/S0165188915000731),”
   *Journal of Economic Dynamics and Control*, 56, 34–54


Stéphane Adjemian's avatar
Stéphane Adjemian committed
541 542 543 544 545 546 547
Announcement for Dynare 4.5.7 (on 2019-02-06)
=============================================

We are pleased to announce the release of Dynare 4.5.7.

This is a bugfix release.

548
The Windows packages are already available for download at: <http://www.dynare.org/download/dynare-stable>.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581

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

This release is compatible with MATLAB versions 7.5 (R2007b) to 9.4 (R2018b)
and with GNU Octave versions 4.4.1.

Here is a list of the problems identified in version 4.5.6 and that have been
fixed in version 4.5.7:

 - The mex-file conducting the QZ decomposition erroneously applied
   the `qz_criterium` to the square absolute value of eigenvalues
   instead of the absolute value itself (as done in mjdgges.m and the
   AIM solver).

 - In pathological cases, `mode_compute=5` (`newrat`) might enter an
   infinite loop.

 - `discretionary_policy` might erroneously state that the derivatives
   of the objective function are non-zero if there are NaN present.

 - Dynare++, when conducting the QZ decomposition, erroneously applied
   the `qz_criterium` to the square absolute value of eigenvalues
   instead of the absolute value itself.

 - Dynare++: IRFs were incorrectly computed.

 - `dynare_sensitivity` did not display the figures of
   `irf_calibration`, it only stored them on the disk.

 - Scatter plots generated by `dynare_sensitivity` did not correctly
   display LaTeX names.

 - Parameter updating via steady state files did not correctly work in
582
   case of using `[static]`/`[dynamic]` equation tags.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603

 - Memory leaks in `k_order_pert` (used by higher order stochastic
   simulations) could lead to crashes.

 - Predetermined variables were not properly set when used in model
   local variables.

 - Posterior moment computation did not correctly update the
   covariance matrix of exogenous shocks during posterior sampling.

 - Dynare was crashing with a cryptic message if a non estimated
   parameter was initialized in the `estimated_params_init` block.

 - The `forecast` command crashed if the model was declared as linear
   and contained deterministic exogenous variables.

 - Block decomposition is broken when used in conjunction with
   `varexo_det`.

 - The model was not correctly specified when `identification` was run
   without another stochastic command in the `.mod` file
604
   (*e.g.* `estimation`, `stoch_simul`, etc.).
Stéphane Adjemian's avatar
Stéphane Adjemian committed
605 606 607 608 609 610 611 612 613 614 615

 - Realtime annualized shock decompositions added the wrong steady state
   value.

 - `mh_recover` option crashed when using slice sampler.

 - x-axis values in plots of moment restrictions were wrong for
   autocovariances.



616 617 618 619 620 621 622 623
Announcement for Dynare 4.5.6 (on 2018-07-25)
=============================================

We are pleased to announce the release of Dynare 4.5.6.

This is a bugfix release.

The Windows packages are already available for download at:
624
<http://www.dynare.org/download/dynare-stable>.
625 626 627

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

628
This release is compatible with MATLAB versions 7.5 (R2007b) to 9.4 (R2018a)
629 630 631 632 633 634 635 636 637
and with GNU Octave versions 4.4.

Here is a list of the problems identified in version 4.5.5 and that have been
fixed in version 4.5.6:

 - TaRB sampler: incorrect last posterior was returned if the last draw was
   rejected.

 - Fixed online particle filter by drawing initial conditions in the prior
Stéphane Adjemian's avatar
Stéphane Adjemian committed
638
   distribution.
639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660

 - Fixed evaluation of the likelihood in non linear / particle filters.

 - Added missing documented `montecarlo` option in Gaussian Filter and
   Nonlinear Kalman Filter.

 - Added back a flag to deal with errors on Cholesky decomposition in the
   Conditional Particle Filter.

 - Macroprocessor `length()` operator was returning 1 when applied to a
   string. Macroprocessor now raises an error when `length()` operator is
   called on an integer and return the number of characters when applied to a
   string.

 - `mode_compute=8`: the error code during mode-finding was not correctly
   handled, resulting in crashes.

 - Identification was not correctly displaying a message for collinear parameters
   if there was no unidentified parameter present.



Stéphane Adjemian's avatar
Stéphane Adjemian committed
661 662 663 664 665 666 667 668
Announcement for Dynare 4.5.5 (on 2018-06-08)
=============================================

We are pleased to announce the release of Dynare 4.5.5.

This is a bugfix release.

The Windows packages are already available for download at:
669
<http://www.dynare.org/download/dynare-stable>.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
670 671 672

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

673
This release is compatible with MATLAB versions 7.5 (R2007b) to 9.4 (R2018a)
Stéphane Adjemian's avatar
Stéphane Adjemian committed
674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699
and with GNU Octave versions 4.2.

Here is a list of the problems identified in version 4.5.4 and that have been
fixed in version 4.5.5:

 - Identification was crashing during prior sampling if `ar` was initially too
   low.

 - The `align` method on `dseries` did not return a functional second `dseries`
   output.

 - Predetermined variables were not properly set when used in model local
   variables.

 - `perfect_foresight_solver` with option `stack_solve_algo=7` was not working
   correctly when an exogenous variable has a lag greater than 1.

 - `identification` with `prior_mc` option would crash if the number of moments
   with non-zero derivative is smaller than the number of parameters.

 - Calling several times `normcdf` or `normpdf` with the same arguments in a
   model with block decomposition (but not bytecode) was leading to incorrect
   results.



700 701 702 703 704 705 706 707
Announcement for Dynare 4.5.4 (on 2018-01-29)
=============================================

We are pleased to announce the release of Dynare 4.5.4.

This is a bugfix release.

The Windows packages are already available for download at:
708
<http://www.dynare.org/download/dynare-stable>.
709 710 711

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

712
This release is compatible with MATLAB versions 7.5 (R2007b) to 9.3 (R2017b)
713 714 715 716 717 718 719 720 721 722 723 724 725 726 727
and with GNU Octave versions 4.2.

Here is a list of the problems identified in version 4.5.3 and that have been
fixed in version 4.5.4:

 - The `type` option of `plot_shock_decomposition` was always set to `qoq` regardless of what is specified.

 - Bug in GSA when no parameter was detected below pvalue threshold.

 - Various bug fixes in shock decompositions.

 - Bug in reading in macro arrays passed on `dynare` command line via the `-D` option.

 - Estimation with missing values was crashing if the `prefilter` option was used.

728
 - Added a workaround for a difference in behaviour between Octave and MATLAB regarding the creation
729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747
   of function handles for functions that do not exist in the path. With Octave 4.2.1, steady state
   files did not work if no auxiliary variables were created.

 - The `stoch_simul` command was crashing with a cryptic message if option `order=3` was used without
   setting `k_order_solver`.

 - In cases where the prior bounds are infinite and the mode is estimated at exactly 0, no `mode_check`
   graphs were displayed.

 - Parallel execution of MCMC was broken in models without auxiliary variables.

 - Reading data with column names from Excel might crash.

 - The multivariate Kalman smoother was crashing in case of missing data in the observations and
   `Finf` became singular.

 - The `plot_shock_decomposition` command ignored various user-defined options like `fig_name`,
   `use_shock_groups` or `interactive` and instead used the default options.

748
 - Nested `@#ifdef` and `@#ifndef` statements don’t work in the macroprocessor.
749 750 751



752 753 754 755 756 757 758 759 760
Announcement for Dynare 4.5.3 (on 2017-10-19)
=============================================

We are pleased to announce the release of Dynare 4.5.3.

This is a bugfix release. It comes less than 24 hours after the previous release,
because version 4.5.2 was affected by a critical bug for MATLAB older than R2016b.

The Windows packages are already available for download at:
761
<http://www.dynare.org/download/dynare-stable>.
762 763 764

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

765
This release is compatible with MATLAB versions 7.5 (R2007b) to 9.3 (R2017b)
766 767
and with GNU Octave versions 4.2.

Stéphane Adjemian's avatar
Stéphane Adjemian committed
768 769
Here is a list of the problems identified in version 4.5.2 and that have been
fixed in version 4.5.3:
770 771


772
 - `isfile` routine was failing with MATLAB older than R2016b. This bug did not
773 774 775 776
   affect Octave.



777 778 779 780 781 782 783 784
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:
785
<http://www.dynare.org/download/dynare-stable>.
786 787 788

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

789
This release is compatible with MATLAB versions 7.5 (R2007b) to 9.3 (R2017b)
790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836
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
Stéphane Adjemian's avatar
Stéphane Adjemian committed
837
   getting parameters. As a consequense, the `shock_decomposition`-command
838 839 840 841 842 843 844
   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.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
845

846 847 848 849 850
 - 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
851
   MATLAB/Windows 7).
852 853 854



855 856 857 858 859 860 861 862
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:
863
<http://www.dynare.org/download/dynare-stable>.
864 865 866

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

867
This release is compatible with MATLAB versions 7.5 (R2007b) to 9.2 (R2017a)
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
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.



898
Announcement for Dynare 4.5.0 (on 2017-06-11)
Stéphane Adjemian's avatar
Stéphane Adjemian committed
899 900 901 902 903 904 905
=============================================

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:
906
<http://www.dynare.org/download/dynare-stable>.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
907 908 909 910 911

The Mac and Debian/Ubuntu packages should follow soon.

All users are strongly encouraged to upgrade.

912
This release is compatible with MATLAB versions ranging from 7.5 (R2007b) to
Stéphane Adjemian's avatar
Stéphane Adjemian committed
913 914 915 916 917 918 919
9.2 (R2017a) and with GNU Octave version 4.2.

Here is the list of major user-visible changes:


 - Ramsey policy

920
   + Added command `ramsey_model` that builds the expanded model with
921
     FOC conditions for the planner’s problem but doesn’t perform any
922 923
     computation. Usefull to compute Ramsey policy in a perfect
     foresight model,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
924

925 926
   + `ramsey_policy` accepts multipliers in its variable list and
     displays results for them.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
927 928 929 930


 - Perfect foresight models

931 932 933 934 935 936 937 938
   + 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
939

940 941 942
   + 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
943

944 945
   + New option `stack_solve_algo=7` that allows specifying a
     `solve_algo` solver for solving the model,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
946

947 948
   + 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
949

950 951
   + New option `lmmcp` that solves the model via a Levenberg-Marquardt
     mixed complementarity problem (LMMCP) solver,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
952

953 954
   + 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
955

956 957
   + New options `tolf` and `tolx` to control termination criteria of
     solvers,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
958

959
   + New option `endogenous_terminal_period` to `simul`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
960

961 962
   + 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
963 964 965 966


 - Optimal simple rules

967
   + Saves the optimal value of parameters to `oo_.osr.optim_params`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
968

969 970
   + New block `osr_params_bounds` allows specifying bounds for the
     estimated parameters,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
971

972 973
   + 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
974

975 976 977
   + 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
978 979 980 981


 - Forecasts and Smoothing

982 983
   + The smoother and forecasts take uncertainty about trends and means
     into account,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
984

985 986
   + 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
987

988 989
   + 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
990

991 992
   + new field `oo_.Smoother.TrendCoeffs` that stores the trend
     coefficients.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
993

994 995
   + Rolling window forecasts allowed in `estimation` command by
     passing a vector to `first_obs`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
996

997 998
   + The `calib_smoother` command now accepts the `loglinear`,
     `prefilter`, `first_obs` and `filter_decomposition` options.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
999 1000 1001 1002


 - Estimation

1003 1004 1005
   + 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
1006

1007
   + `load_mh_file` and `mh_recover` now try to load chain’s proposal density,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1008

1009 1010 1011
   + 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
1012

1013 1014 1015
   + 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
1016

1017 1018
   + Saves the posterior density at the mode in
     `oo_.posterior.optimization.log_density`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1019

1020 1021
   + The `filter_covariance` option now also works with posterior
     sampling like Metropolis-Hastings,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1022

1023 1024
   + New option `no_posterior_kernel_density` to suppress computation
     of kernel density of posterior objects,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1025

1026 1027
   + Recursive estimation and forecasting now provides the individual
     `oo_` structures for each sample in `oo_recursive_`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1028

1029
   + The `trace_plot` command can now plot the posterior density,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1030

1031 1032
   + New command `generate_trace_plots` allows generating all trace
     plots for one chain,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1033

1034 1035 1036
   + 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
1037

1038 1039
   + New option `huge_number` for replacement of infinite bounds with
     large number during `mode_compute`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1040

1041 1042 1043 1044 1045 1046
   + 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
1047

1048 1049 1050 1051
   + 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
1052

1053 1054
   + New option `raftery_lewis_diagnostics` that computes *Raftery and Lewis
     (1992)* convergence diagnostics,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1055

1056
   + New option `fast_kalman_filter` that provides fast Kalman filter
1057
     using Chandrasekhar recursions as described in *Ed Herbst (2015)*,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1058

1059 1060
   + The `dsge_var` option now saves results at the posterior mode into
     `oo_.dsge_var`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1061

1062 1063
   + 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
1064

1065
   + New prior density: generalized Weibull distribution,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1066

1067 1068
   + Option `mh_recover` now allows continuing a crashed chain at the
     last save mh-file,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1069

1070 1071 1072
   + 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
1073

1074 1075 1076
   + 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
1077

1078 1079
   + The `use_calibration` to `estimated_params_init` now also works
     with ML,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1080

1081
   + Improved initial estimation checks.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1082 1083 1084 1085


 - Steady state

1086 1087 1088
   + 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
1089

1090 1091
   + New options `tolf` and `tolx` to control termination criteria of
     solver,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1092

1093 1094
   + The debugging mode now provides the termination values in steady
     state finding.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1095 1096 1097 1098


 - Stochastic simulations

1099
   + New options `nodecomposition`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1100

1101 1102
   + New option `bandpass_filter` to compute bandpass-filtered
     theoretical and simulated moments,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1103

1104 1105
   + New option `one_sided_hp_filter` to compute one-sided HP-filtered
     simulated moments,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1106

1107 1108
   + `stoch_simul` displays a simulated variance decomposition when
     simulated moments are requested,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1109

1110 1111
   + `stoch_simul` saves skewness and kurtosis into respective fields
     of `oo_` when simulated moments have been requested,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1112

1113 1114
   + `stoch_simul` saves the unconditional variance decomposition in
     `oo_.variance_decomposition`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1115

1116 1117
   + New option `dr_display_tol` that governs omission of small terms
     in display of decision rules,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1118

1119 1120
   + 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
1121

1122 1123
   + The `loglinear` option now works with lagged and leaded exogenous
     variables like news shocks,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1124

1125 1126
   + New option `spectral_density` that allows displaying the spectral
     density of (filtered) endogenous variables,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1127

1128 1129
   + New option `contemporaneous_correlation` that allows saving
     contemporaneous correlations in addition to the covariances.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1130 1131 1132 1133


 - Identification

1134
   + New options `diffuse_filter` and `prior_trunc`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1135

1136 1137
   + The `identification` command now supports correlations via
     simulated moments,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1138 1139 1140 1141


 - Sensitivity analysis

1142
   + New blocks `irf_calibration` and `moment_calibration`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1143

1144
   + Outputs LaTeX tables if the new `TeX` option is used,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1145

1146
   + New option `relative_irf` to `irf_calibration` block.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1147 1148 1149 1150


 - Conditional forecast

1151 1152
   + Command `conditional_forecast` now takes into account `histval`
     block if present.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1153 1154 1155 1156


 - Shock decomposition

1157 1158
   + New option `colormap` to `shocks_decomposition` for controlling
     the color map used in the shocks decomposition graphs,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1159

1160
   + `shocks_decomposition` now accepts the `nograph` option,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1161

1162 1163
   + New command `realtime_shock_decomposition` that for each period `T= [presample,...,nobs]`
     allows computing the:
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1164

1165
     * realtime historical shock decomposition `Y(t|T)`, *i.e.* without observing data in `[T+1,...,nobs]`
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1166

1167
     * forecast shock decomposition `Y(T+k|T)`
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1168

1169 1170 1171 1172 1173 1174 1175 1176 1177
     * 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
1178 1179 1180 1181


 - Macroprocessor

1182
   + Can now pass a macro-variable to the `@#include` macro directive,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1183

1184 1185 1186
   + 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
1187 1188 1189 1190


 - Command line

1191
   + New option `onlyclearglobals` (do not clear JIT compiled functions
1192
     with recent versions of MATLAB),
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1193

1194 1195
   + New option `minimal_workspace` to use fewer variables in the
     current workspace,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1196

1197 1198 1199
   + 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
1200

1201 1202
   + 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
1203 1204 1205 1206


 - dates/dseries/reporting classes

1207
   + New methods `abs`, `cumprod` and `chain`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1208

1209
   + New option `tableRowIndent` to `addTable`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1210

1211 1212
   + Reporting system revamped and made more efficient, dependency on
     matlab2tikz has been dropped.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1213 1214 1215 1216


 - Optimization algorithms

1217
   + `mode_compute=2` Uses the simulated annealing as described by
1218
     *Corana et al. (1987)*,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1219

1220 1221
   + `mode_compute=101` Uses SOLVEOPT as described by *Kuntsevich and
     Kappel (1997)*,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1222

1223
   + `mode_compute=102` Uses `simulannealbnd` from MATLAB’s Global
1224
     Optimization Toolbox (if available),
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1225

1226 1227
   + New option `silent_optimizer` to shut off output from mode
     computing/optimization,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1228

1229 1230
   + New options `verbosity` and `SaveFiles` to control output and
     saving of files during mode computing/optimization.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1231 1232 1233 1234


 - LaTeX output

1235
   + New command `write_latex_original_model`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1236

1237 1238 1239
   + 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
1240

1241 1242
   + 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
1243

1244 1245
   + 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
1246

1247 1248
   + New command `collect_latex_files` that creates one compilable LaTeX
     file containing all TeX-output.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1249 1250 1251 1252


 - Misc.

1253
   + Provides 64bit preprocessor,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1254

1255 1256
   + Introduces new path management to avoid conflicts with other
     toolboxes,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1257

1258
   + Full compatibility with MATLAB 2014b’s new graphic interface,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1259

1260 1261
   + When using `model(linear)`, Dynare automatically checks
     whether the model is truly linear,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1262

1263 1264
   + `usedll`, the `msvc` option now supports `normcdf`, `acosh`,
     `asinh`, and `atanh`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1265

1266 1267 1268
   + 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
1269

1270 1271
   + Improved numerical performance of
     `schur_statespace_transformation` for very large models,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1272

1273
   + The `all_values_required` option now also works with `histval`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1274

1275
   + Add missing `horizon` option to `ms_forecast`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1276

1277 1278 1279 1280
   + 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
1281 1282 1283



1284
Bugs and problems identified in version 4.4.3 and that have been fixed in version 4.5.0:
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1285 1286 1287 1288


 - BVAR models

1289 1290
   + `bvar_irf` could display IRFs in an unreadable way when they moved from
     negative to positive values,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1291

1292 1293
   + 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
1294 1295 1296 1297


 - Conditional forecasts

1298 1299 1300
   + 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
1301

1302 1303
   + The `plot_conditional_forecast` option could produce unreadable figures if
     the areas overlap,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1304

1305
   + The `conditional_forecast` command after MLE crashed,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1306

1307 1308
   + 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
1309

1310 1311 1312
   + 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
1313 1314 1315 1316


 - Discretionary policy

1317 1318
   + Dynare allowed running models where the number of instruments did not match
     the number of omitted equations,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1319

1320
   + Dynare could crash in some cases when trying to display the solution,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1321

1322 1323
   + Parameter dependence embedded via a `steady_state` was not taken into
     account, typically resulting in crashes.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1324 1325 1326

 - dseries class

1327 1328
   + When subtracting a dseries object from a number, the number was instead
     subtracted from the dseries object.
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1329 1330 1331 1332


 - DSGE-VAR models

1333 1334
   + Dynare crashed when estimation encountered non-finite values in the Jacobian
     at the steady state,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1335

1336 1337 1338 1339
   + 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
1340 1341 1342 1343


 - Estimation command

1344 1345
   + 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
1346

1347
   + Calling estimation after identification could lead to crashes,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1348

1349 1350 1351 1352
   + 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
1353

1354 1355
   + Computation of Bayesian smoother could crash for larger models when
     requesting `forecast` or `filtered_variables`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1356

1357 1358
   + 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
1359

1360 1361
   + The Geweke convergence diagnostics always used the default `taper_steps` and
   `geweke_interval`,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1362

1363 1364
   + 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
1365

1366 1367
   + If `bayesian_irfs` was requested when `mh_replic` was too low to compute
     HPDIs, plotting was crashing,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1368

1369 1370 1371
   + 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
1372

1373
   + Using a user-defined `mode_compute` crashed estimation,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1374

1375
   + Option `mode_compute=10` did not work with infinite prior bounds,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1376

1377 1378 1379
   + 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
1380

1381 1382
   + Using the `forecast` option with `loglinear` erroneously added the unlogged
     steady state,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1383

1384 1385
   + 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
1386

1387 1388
   + 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
1389

1390 1391 1392
   + 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
1393

1394 1395 1396
   + 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
1397

1398 1399
   + 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
1400

1401 1402 1403
   + 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
1404

1405 1406 1407
   + 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
1408

1409
   + The estimation option `lyapunov=fixed_point` was broken,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1410

1411
   + Computation of `filtered_vars` with only one requested step crashed Dynare,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1412

1413
   + Option `kalman_algo=3` was broken with non-diagonal measurement error,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1414

1415
   + When using the diffuse Kalman filter with missing observations, an additive
1416
     factor log(2π) was missing in the last iteration step,
Stéphane Adjemian's avatar
Stéphane Adjemian committed
1417