Commit Graph

314 Commits (bbd12a226e4108202a3505fd7c96d35370286c45)

Author SHA1 Message Date
Johannes Pfeifer 0f50f33c2c Initialize M_.dname=M_.fname in global_initialization.m
Closes #979
2015-07-28 15:23:53 +02:00
Michel Juillard ef8f36565b initializing options_.instruments = [] 2015-07-27 15:52:14 +02:00
Michel Juillard 9b8077fc28 Merge branch 'master' into new_ep 2015-07-21 09:30:58 +02:00
Stéphane Adjemian (Charybdis) 8007f508a2 New option linear_approximation for perfect foresight models.
This approach only requires one evaluation of the dynamic model (and its
jacobian) instead of T (the number of perdiods). Also (because the model
is linear) the equilibrium paths are obtained by inverting the jacobian
of the stacked equations (no need for a Newton algorithm).

Only available with stack_solve_algo==0 (which is the default algorithm
for solving perfect foresight models).

If possible, the option is triggered automatically if the model is
declared linear.

TODO:
 * Write a linear version of perfect_foresight_problem routine.
 * Evaluate the approxilation error (just need to evaluate the system of
 stacked non linear equations).
2015-07-07 17:55:41 +02:00
Johannes Pfeifer 3bf13dd53b Add functionality for TaRB 2015-06-08 16:44:44 +02:00
Johannes Pfeifer 0769fd1cce Add verbosity and SaveFile options to some optimizers
Also moves additional optimizer-related files to corresponding subfolder
Closes #894
2015-06-08 16:44:44 +02:00
Stéphane Adjemian (Hermes) 959349d30f Put rgwmh options in a substructure. 2015-05-28 18:03:53 +02:00
Stéphane Adjemian (Hermes) 194bdb57c3 Merge branch 'rewrite-nonlinear-filters' 2015-05-28 15:40:39 +02:00
Houtan Bastani fe8750c1bc preprocessor: add minimal_workspace option to dynare statement, #946 2015-05-28 11:50:39 +02:00
Frédéric Karamé 0c2483278c Add three variables for RWGMH. 2015-05-28 11:45:16 +02:00
Frédéric Karamé 38a276fb08 restore the default value for distribution_approximation 2015-05-27 14:57:01 +02:00
Michel Juillard b3047c9742 Merge branch 'master' into new_ep 2015-05-23 18:26:07 +02:00
Houtan Bastani e5df26a6e8 preprocessor: add use_tarb option to estimation. #940 2015-05-13 15:32:24 +02:00
Stéphane Adjemian (Hermes) f960e7bb68 Merged branch simpsa-inf-fix. 2015-05-07 16:01:43 +02:00
Stéphane Adjemian (Hermes) 6d0dfc2d3f Do not use hard coded value for the finite interval replacing unbounded intervals in simpsa and simulated annealing algorithms. 2015-05-07 15:59:46 +02:00
Stéphane Adjemian de526eae57 Merge pull request #906 from JohannesPfeifer/table_display
Cosmetic fixes to the on-screen display of tables
2015-05-07 15:15:38 +02:00
Stéphane Adjemian efc5ec7002 Merge pull request #908 from JohannesPfeifer/optimizer_doc
Add documentation for new optimizers
2015-05-07 15:12:15 +02:00
Houtan Bastani 18e9521450 preprocessor: add dirname option to estimation. closes #910 2015-04-27 11:59:21 +02:00
Johannes Pfeifer 9408319e48 Cosmetic changes related to solvopt.m 2015-04-24 12:31:27 +02:00
Johannes Pfeifer 8285121f9b Allow selection of tolerance for display of decision rules 2015-04-22 15:37:35 +02:00
Stéphane Adjemian (Charybdis) 1970951d1f Merge branch 'johannes_fix_simulated_annealing_and_other_optimization_related_issues' 2015-04-16 22:31:07 +02:00
Marco Ratto fe48de4406 Re-set default of new option diffuse_kalman_tol to 1e-6:
former hardcoded value was 1e-6 in the univariate smoother (i.e. the one most often used).
Updated documentation.
2015-04-08 15:44:39 +02:00
Johannes Pfeifer a81c9da9a2 Add simulated_annealing.m to optimizers
Restores compatibility of mode_compute=102
2015-04-06 11:08:20 +02:00
Johannes Pfeifer 405b1f7368 Add solvopt.m to Dynare optimizers
Restores compatibility of mode_compute=101
2015-04-06 11:05:12 +02:00
Stéphane Adjemian (Charybdis) 090c4fedbd Added new option (diffuse_kalman_tol) and fixed tolerance paremeters in diffuse smoother routines. 2015-04-03 18:02:03 +02:00
Marco Ratto c881cfff17 Fixed issues related to newratflag.
This options relates to alternative numerical hessian computations:

optim=('Hessian',1) is the default dynare numeric Hessian
optim=('Hessian',0) is the outer product gradient
optim=('Hessian',2) is the 'mixed' outer product gradient, where diagonal elements using second order derivation formula,

Both 0 and 2 cases require univariate filters, to ensure using maximum number of individual densities.
2015-03-12 17:24:21 +01:00
Houtan Bastani 6db9f10353 preprocessor: remove use of tables completely as they were introduced in Matlab R2013b. #824 2015-03-09 17:59:05 +01:00
Houtan Bastani d77b5104d2 preprocessor: cell2table doesn’t exist in Octave 2015-03-09 17:16:59 +01:00
Houtan Bastani 30428aeb17 preprocessor: add joint prior syntax, #824 2015-03-03 15:08:33 +01:00
Johannes Pfeifer 0ef7524977 Use dynare_minimize_objective for OSR computations 2015-02-28 20:37:10 +01:00
Stéphane Adjemian (Charybdis) 0bb413e8fa Merge branch 'experimental-optimizers' 2015-02-24 20:44:39 +01:00
Stéphane Adjemian (Charybdis) 5ef340b8cf Rewrote gmhmaxlik routine (mode_compute==6). 2015-02-24 20:34:35 +01:00
Michel Juillard 21f009a531 Merge branch 'master' into new_ep 2015-02-14 11:54:57 +01:00
Houtan Bastani f72910ae33 fix preprocessor implimentation of filter_algorithm option to estimation. closes #843 2015-02-09 03:20:37 +01:00
Michel Juillard 9c6e219990 new implementation for extended path 2015-02-06 12:36:09 +01:00
Johannes Pfeifer dd6f8c182c Move call to optimizers in mode-computing to separate file to allow other function to access all integrated optimizers
- also moves several options to subfields of options structure
- allows setting options of newrat
- requires newrat to have compatible kalman_algo specified in order for it to not change options_
- explicitly disallows using analytical derivation with numerical gradient (before the numerical gradient request was overwritten)
- always outputs hessian returned by optimizer (empty matrix if not computed) and deletes subsequent overwriting if cova_compute=0
2014-12-04 19:36:12 +01:00
Johannes Pfeifer 268e8627f0 Add missing simple option to global_initialization.m 2014-11-25 09:54:41 +01:00
Marco Ratto 06572f26a4 Use the nodecomposition option which is available in th_autocovariances.m: for large models with the HP filter option, this may spare a lot of time, if only the unconditional moments are needed on output; 2014-11-14 17:28:17 +01:00
Johannes Pfeifer cc96d46911 Transfer hard-coded verbosity of simplex algorithm into an option 2014-11-10 19:54:44 +01:00
Johannes Pfeifer a9bb341b0a Increase default maxit for steady and deterministic simulations
Judging from forum reports, the default fails too often. Closes #747
2014-09-30 11:47:58 +02:00
Johannes Pfeifer 8655df67a9 Clean up use of verbosity option
- in dr_block.m and ep_residuals.m the option was hardcoded and disabled
- for csminwel1.m, no verbosity option exists anymore, thus definitions preceeding calls to it were redundant as was the options_.osr.verbose
2014-09-12 10:33:37 +02:00
Stéphane Adjemian (Scylla) 5500846280 Added option distribution_proposal (estimation, non linear filters). 2014-09-08 22:36:41 +02:00
Stéphane Adjemian (Scylla) 4c9be8a56a Removed useless option. 2014-09-08 22:36:41 +02:00
Stéphane Adjemian (Scylla) df19396acf Added option proposal_approximation (estimation, non linear filters). 2014-09-08 22:36:41 +02:00
Stéphane Adjemian (Scylla) 9bfdeb30e2 Bug fixes. 2014-09-08 22:36:41 +02:00
Stéphane Adjemian (Scylla) 5e4b7d8d42 Added filter_algorithm (estimation command, sets the particle filter algorithm). 2014-09-08 22:36:41 +02:00
Stéphane Adjemian (Scylla) bcad4f31f2 Added option resampling_method (estimation command, non linear filters). 2014-09-08 22:36:41 +02:00
Stéphane Adjemian (Scylla) 08c74b8f41 Added option resampling_threshold (estimation command, non linear filters). 2014-09-08 22:36:40 +02:00
Stéphane Adjemian (Scylla) a6009908e4 Added option resampling (estimation command, non linear filters). 2014-09-08 22:36:40 +02:00
Stéphane Adjemian (Scylla) f6a1df8549 Increased default number of particles. 2014-09-08 22:36:40 +02:00