9408319e48
Cosmetic changes related to solvopt.m
2015-04-24 12:31:27 +02:00
4ffebeee2e
dseries: submodule update
2015-04-23 14:15:47 +02:00
db272f114b
Compatibility fixes for MATLAB <= 7.7.
...
The tilde syntax was introduced in MATLAB 7.8.
2015-04-22 16:45:15 +02:00
8285121f9b
Allow selection of tolerance for display of decision rules
2015-04-22 15:37:35 +02:00
20a4f551ed
Improve display of decision rules by accounting for variable lengths
...
Also cleans up file by factoring code
2015-04-22 15:29:04 +02:00
0dac179944
Improve display of tables
...
Closes #892 by allowing for optional inputs for cell lengths. Replaces centering of headers by right-flush to improve readability.
2015-04-22 15:26:03 +02:00
f86b91d3d6
Perfect foresight solver: inform the user when we switch to a homotopy method.
2015-04-22 14:44:39 +02:00
ac15eda6de
Revert "Add check for valid filename"
...
This reverts commit 39c7b793f29f7d3524e9
2015-04-22 13:55:35 +02:00
57f89d6843
Merge pull request #905 from JohannesPfeifer/filename_check
...
Add check for valid filename
2015-04-22 13:47:15 +02:00
9f7d3524e9
Add check for valid filename
...
Uses the same restrictions as variable names.
2015-04-22 13:30:09 +02:00
8ab1239d79
Fix handling of warnings in perfect foresight solver with homotopy.
2015-04-22 12:23:20 +02:00
df697febcb
dseries: submodule update
2015-04-22 11:38:36 +02:00
8eaa645800
Fix compatibility of mode_compute=7
...
Uses wrapper as for mode_compute=3 to restore to conform to Octave syntax
2015-04-22 11:08:32 +02:00
39c7b793f2
Add check for valid filename
...
Uses the same restrictions as variable names.
2015-04-22 10:57:10 +02:00
c6c5cd8ce8
Allow Inf bounds in simpsa by replacing them with big finite number
...
Closes #585
2015-04-21 16:26:46 +02:00
3dee87c243
Fix crash in unit test fs2000_smoother_only
...
For calibrated models, the param_vals field does not exist
2015-04-20 16:45:51 +02:00
31cd80e9a9
Fix crash in error message display of perfect_foresight_setup.m when more than one variable is not present in current period
...
Concatenation otherwise crashes due to non-conformable matrices
2015-04-19 10:52:22 +02:00
8ab8d46281
small fix in matlab annotation
2015-04-17 09:16:54 +02:00
0a27b71f7e
Merge pull request #886 from rattoma/gsa
...
Enhance and harmonize reduced form mapping routine
2015-04-17 08:41:32 +02:00
bc6cf0345f
Clean up csminwel1.m
...
Adds documentation and puts calling of numerical gradient to nested function as previously suggested in comment at the bottom of file
2015-04-17 07:05:01 +02:00
b629d3ca19
Cosmetic changes to csminwel routines
...
Mostly adds documentation
2015-04-17 07:05:00 +02:00
84f0060cbd
Fix bug where inverse Hessian from csminwel1.m was taken as Hessian
2015-04-17 07:05:00 +02:00
1970951d1f
Merge branch 'johannes_fix_simulated_annealing_and_other_optimization_related_issues'
2015-04-16 22:31:07 +02:00
b0f4f2af78
Harmonize use of front-end mcf_analysis.m in GSA and identification routines.
2015-04-16 18:25:35 +02:00
57e9b9fa8b
Revived reduced form mapping routine:
...
1) use new mcf_analysis.m;
2) use generic filtering when threshold does not work;
3) make log-tranformed mapping meaningful and comparable to baseline;
3) harmonize names of files saved and reduced number of directories created;
2015-04-16 16:13:16 +02:00
0faf6e1981
Allow inverse transform and set minimum threshold for lam
2015-04-16 16:13:00 +02:00
ae6302c348
Added output argument to mcf_analysis.m
2015-04-16 16:09:39 +02:00
4e1e7732fd
fix graphic issues with eps
2015-04-16 12:30:27 +02:00
8d679bff6f
missing bits in fixes to graphical issues #881 .
2015-04-16 08:52:42 +02:00
41cebf7e4e
- fix nograph option in GSA graphical output
...
- fix graphical output issues on unix
2015-04-15 22:54:36 +02:00
f0770186ca
Exhanged mode_compute=2 and mode_compute=102.
...
These options are for the simulated annealing algorithm, the Mathworks' global optimization
toolbox is mandatory for mode_compute=102. The first option relies on a
free implementation (works also with Octave).
2015-04-15 12:42:08 +02:00
da2c529ee7
Merge branch 'marco_gsa_fixes'
2015-04-14 15:54:59 +02:00
1a6afc0fdb
Restore compatibility of recursive estimation with new data interface
...
Use dataset_ structure instead of reconstructing data to plot. Eliminates crashes
2015-04-14 15:12:25 +02:00
0bd8fef202
- Various bug fixes on moment calibration
...
- Bug fixes on plots showing restrictions
- Improved and increased graphical info
2015-04-14 12:24:49 +02:00
ffc4b7480c
- improved plots with MATLAB annotations [trapped in octave] and different colors
...
- use uimenu to get the name of the observable associated to each CDF from MATLAB plots
[trapped in octave]
- also checks multiple significant parameters by individual observable
2015-04-14 12:16:50 +02:00
a44b3fbfe7
Bug fixed for incomplete stock matrices of log-posterior kernel and steady state
...
[used in gsa/filt_mc_.m]
2015-04-14 12:10:46 +02:00
e32cbf626f
Fixed bug when neighborhood_width is used : MC option must be forced.
2015-04-14 12:06:25 +02:00
ca8f0ea006
Harmonize filters/likelihood with smoothers by using new option diffuse_kalman_tol
2015-04-08 15:49:12 +02:00
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
49d4643481
Fix copy and paste typo in gmhmaxlik.m
2015-04-06 12:04:11 +02:00
d1a80acb44
Replace non-available simulated annealing by Matlab's routine
...
Closes #814
2015-04-06 11:28:32 +02:00
39ed3eb3f0
Fix bugs related to allowing mode_compute to be a string
...
Crashes otherwise
2015-04-06 11:24:42 +02:00
a81c9da9a2
Add simulated_annealing.m to optimizers
...
Restores compatibility of mode_compute=102
2015-04-06 11:08:20 +02:00
405b1f7368
Add solvopt.m to Dynare optimizers
...
Restores compatibility of mode_compute=101
2015-04-06 11:05:12 +02:00
23b322d6c6
Merge pull request #866 from JohannesPfeifer/mode_check
...
Remove hard-coded coloring of legend in mode_check.m
2015-04-03 18:07:14 +02:00
6aa5ceabb2
Fixed bug introduced in 363c4a61222103fda22adff9bdb1115a6073428a.
2015-04-03 18:02:03 +02:00
090c4fedbd
Added new option (diffuse_kalman_tol) and fixed tolerance paremeters in diffuse smoother routines.
2015-04-03 18:02:03 +02:00
e97e5c3407
Exclude zero columns of T from Kitagawa transformation, since ordschur is extremely noisy for multiple zero eigenvalues.
...
This can make a lot of difference for large models that have hundreds of definitions.
2015-04-03 18:02:03 +02:00
5297836577
Harmonize criteria for exiting diffuse steps in likelihood with the smoother.
...
Since initial Pinf is well scaled to unity, crit1= 1.e-6 is used for smoother and should also apply to likelihood evaluations.
2015-04-03 18:02:03 +02:00
7c838b133f
Use local variables for restrict var list, to avoid overwriting the global and messing likelihood evaluations later on.
2015-04-03 18:02:03 +02:00
8cc9fbd69a
Remove hard-coded coloring of legend in mode_check.m
...
Now displays correct colors with new Matlab plot interface (>2014a)
2015-04-02 20:52:49 +02:00
f9154273d0
Merge pull request #854 from JohannesPfeifer/smoother_var_initialization_fix
...
Fix size of filtered variables matrix
2015-04-01 17:59:12 +02:00
f3d1a6c68e
Merge branch 'master' into marco_analytic_derivation
2015-04-01 16:41:38 +02:00
27403fb2dd
Add check whether initial prior is Inf
...
Requires adding fourth output argument of priordens.m that stores position of problematic parameter.
2015-04-01 16:30:46 +02:00
9c3c0d727f
Add check for point priors and force user to fix this parameter.
...
Traps Inf-prior
2015-04-01 16:29:18 +02:00
b4a8ad8aa4
Fix mex path for win64 bit
2015-04-01 14:43:52 +02:00
214dc74723
- Fixed bugs around analytic derivation.
...
- Fixed test routine, eliminating diffuse filter.
- Trapped incompatibility of diffuse filter with analytic derivation.
2015-04-01 09:00:51 +02:00
28f2e2abd0
Merge pull request #856 from rattoma/optimizer_number_5
...
Fixed issues related to newratflag.
2015-03-30 15:22:05 +02:00
ad0e0b6c1d
Merge pull request #861 from JohannesPfeifer/steady_fix
...
Prevent crash if steady state cannot be computed
2015-03-30 15:17:46 +02:00
0e1dd779b9
Check that Dynare's routines are not overridden by other routines.
...
By default Dynare will temporarly change Matlab/Octave's path. Added an
option to force Dynare to keep the current path. Also check that the user
did not put dynare/matlab directory with all the subfolders in the path.
With this patch, we do not need to prefix Dynare routines with dyn_ or
dynare_ anymore.
2015-03-30 15:04:59 +02:00
34c871b53c
getenv provides one output on windows
2015-03-27 18:58:35 +01:00
90e00b3871
fixes for windows
2015-03-27 17:29:10 +01:00
d7c253ae59
support 64-bit preprocessor, #862
2015-03-26 19:08:30 +01:00
781e280610
Prevent crash if steady state cannot be computed
...
If e.g. the log of the steady state cannot be computed, the field oo_.dr.ys does not exist
2015-03-25 09:42:59 +01:00
80f512b90b
ilu function has been implemented in Octave 4.0.
2015-03-18 16:03:41 +01:00
aa1acec308
Remove obsolete warning code for Octave < 3.6.
2015-03-17 18:01:36 +01:00
ec65f80617
Octave warning "matlab-incompatible" renamed to "language-extension" in 4.0.
2015-03-17 17:59:52 +01:00
1a3d8d0b26
Provisions for MATLAB 8.5 (R2015a)
2015-03-16 14:44:06 +01:00
6d7b72cb0d
Merged with master
2015-03-13 15:53:18 +01:00
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
d4c0726026
Fix size of filtered variables matrix
...
Fixes crash occurring due to non-conformable matrices
2015-03-12 10:49:33 +01:00
87cb00bd7d
fixing bug in calling sequence of *_dynamic function when using block option
2015-03-11 09:24:47 +01:00
6e0d919751
Fix issue with purely forward deterministic models with lags on exogenous.
...
Ref #617
2015-03-09 18:22:01 +01:00
6db9f10353
preprocessor: remove use of tables completely as they were introduced in Matlab R2013b. #824
2015-03-09 17:59:05 +01:00
d77b5104d2
preprocessor: cell2table doesn’t exist in Octave
2015-03-09 17:16:59 +01:00
86f9e53ec6
dr_block: fixing bug in solve_foward_complete
2015-03-07 10:59:53 +01:00
265970c1c9
adding global test for identification for sbvar
2015-03-07 10:59:53 +01:00
de5c541ce6
Manually merged PR #807 .
...
Fixed conflict related to the move of the perfect foresight related routines
in a subfolder.
2015-03-05 14:31:36 +01:00
30428aeb17
preprocessor: add joint prior syntax, #824
2015-03-03 15:08:33 +01:00
219b4fe547
Merge pull request #813 from JohannesPfeifer/model_diag
...
Add missing abs to unit root check in model_diagnostics.m
2015-03-01 16:35:31 +01:00
511d621818
Merge pull request #851 from JohannesPfeifer/Estimation_Bugfixes
...
Estimation bugfixes
2015-03-01 16:33:51 +01:00
ba2c4db88c
Merge pull request #816 from JohannesPfeifer/loglinear_steady_error
...
Integrate error from negative steady state with loglinear model into pri...
2015-03-01 16:30:29 +01:00
a7f380d8c0
Condition check for changed parameters on whether parameters are estimated in the first place
...
Fixes crash in master unit test
2015-03-01 13:36:21 +01:00
fe938c902a
Fixed another bug related to options_.optim_opt being now initialized as an empty array
...
Finalizes 392486273e
2015-03-01 13:34:39 +01:00
0ef7524977
Use dynare_minimize_objective for OSR computations
2015-02-28 20:37:10 +01:00
198a6ed7cf
Removed three unused routines.
2015-02-27 18:54:12 +01:00
73cdaa3091
Replaced get_prior_info routine by prior routine.
...
This new routine gives the possibility to obtain a description of the
prior from the command line.
>> prior table
will display a table with the prior mean, prior mode, prior std, prior
bounds and prior HPD interval.
>> prior plot
will plot the prior densities.
>> prior optimize
will trigger the maximization of the prior density and display the
results (often, because of BK conditions or other issues, Dynare is even
not able to get the prior mode).
>> prior moments
will display the moments of the endogenous variables at the prior mode.
>> prior simulate
will run a Monte-Carlo, by sampling from the prior and discarding vector
of parameters such that the steady state does not exist or the BK
conditions are not met, and return an estimate of the effective prior
mass.
2015-02-27 17:12:04 +01:00
daaed37782
Added new routines.
2015-02-27 17:05:11 +01:00
eec1e77140
Removed last character if string is renurned in a variable.
2015-02-26 16:18:33 +01:00
4979f33f8b
Added the possibility to return the generated line in a variable.
2015-02-26 16:10:44 +01:00
04ae10434c
Beta prior. Do not return the prior mean of the prior density has two
...
modes (return the first mode).
2015-02-26 12:07:40 +01:00
392486273e
Fixed bug (options_.optim_opt is now initialized as an empty array).
2015-02-26 08:43:30 +01:00
0bb413e8fa
Merge branch 'experimental-optimizers'
2015-02-24 20:44:39 +01:00
83981125c9
Added missing routine and fixed licence header.
2015-02-24 20:42:19 +01:00
5ef340b8cf
Rewrote gmhmaxlik routine (mode_compute==6).
2015-02-24 20:34:35 +01:00
e19f03b420
replace reporting code with submodule. closes #811
2015-02-24 16:38:36 +01:00
fac1d206ab
Use new wrapper for optimization algorithms when maximizing the prior density.
2015-02-23 18:26:34 +01:00
bd39ff34f6
Fixed bug in get_prior_info routine.
...
lb and ub (lower and upper bounds for the priors) are no more fields of
bayestopt_ structure.
2015-02-23 18:24:57 +01:00
0ffb9dd6e2
Changed order of the input arguments in dynare_mimimize_objective.
...
It is more standard to have the name of the routine (defining the
objective to be minimized) as a first argument.
2015-02-23 14:35:41 +01:00
beb7826376
Moved optimization related routines in a common subfolder.
2015-02-23 14:26:07 +01:00
5865d46a74
Fixed bug. Wrong number of returned arguments when calling
...
dynare_minimize_objective (missing options_).
2015-02-21 18:14:15 +01:00
32c29fece5
Merge pull request #842 from JohannesPfeifer/cosmetics
...
Cosmetical changes and additional documentation
2015-02-21 15:42:51 +01:00
7df3856cbf
Fixed homotopy with bytecode for perfect foresight models.
2015-02-21 13:59:53 +01:00
43f2ec1d9d
Fixed display of the homotopy steps when the residuals are complex.
2015-02-21 13:58:52 +01:00
ff7c1f5cda
Cosmetic change.
2015-02-21 13:57:48 +01:00
9bdf75a52c
Fixed bug in the initialization of the homotopy for perfect foresight models.
2015-02-21 13:56:11 +01:00
402a826300
Updated dates submodule.
2015-02-20 15:16:20 +01:00
38b548c4d0
Fixed licence header.
2015-02-20 14:48:25 +01:00
d19592f761
Fixed homotopy for perfect foresight models with block option, cosmetic changes.
2015-02-20 14:48:25 +01:00
1faf755d56
Oops, fixing dseries branch.
2015-02-19 18:26:49 +01:00
e2d7b8ad25
Update dseries.
2015-02-19 18:23:39 +01:00
5768acf17e
Fixed the verbosity of perfect foresight models solvers.
...
With stack_solve_algo={0, 6} the verbosity is controlled with
options_.verbosity instead of options_.no_homotopy.
Also harmonized the printed informations across these two algorithms.
2015-02-18 23:58:37 +01:00
fd0a172637
Remove small imaginary parts in the solution path for perfect foresight
...
models obtained with stack_solve_algo=7.
Also return oo_.deterministic_simulation.status equal to false if the
solution path is not real.
2015-02-18 23:53:55 +01:00
0851d277c4
Use correct tolerance parameter when solving perfect foresight model
...
with stack_solve_algo=7.
2015-02-18 23:52:17 +01:00
01b83d767a
Added a matlab routine that prints a line by replicating a symbol.
2015-02-18 18:31:39 +01:00
51a277424d
Cosmetic changes.
2015-02-18 18:21:45 +01:00
6f0584fe38
fix copyright dates
2015-02-18 03:06:33 +01:00
1c8e1ab913
Merge branch 'remove-globals-from-blocks'
2015-02-17 13:01:17 +01:00
cc94ed63b0
Revert "Temporary fix for perfect foresight model simulation with block option."
...
This reverts commit 8875dbdaa5
2015-02-17 13:01:01 +01:00
a894357655
Removed options_ and oo_ globals from *all* generated dynamic model
...
routines (with block option) and adapted some routines to reflect this change.
2015-02-17 12:57:30 +01:00
83c68c0477
Allowing for models without exogenous variable
2015-02-16 21:30:34 +01:00
8875dbdaa5
Temporary fix for perfect foresight model simulation with block option.
2015-02-16 17:05:37 +01:00
d3275db830
Removed options_ and oo_ globals from generated dynamic model
...
routines (with block option).
2015-02-16 16:56:17 +01:00
333f87ced2
fix headers
2015-02-16 09:08:26 +01:00
ca83397d95
update copyright dates for files changed in 2015
2015-02-16 09:08:02 +01:00
dfc89df6a8
added make_ex_ to fix bug in dyn_forecast and fixed length of output
2015-02-15 17:49:01 +01:00
b3ed7901cf
Fixed bug in homotopy for perfect foresight models.
...
Homotopy was not working if nans or complex numbers were present in oo_.endo_simul.
2015-02-13 16:04:25 +01:00
3b41aa0e60
Fixed bug related to the type of initial_period.
2015-02-13 15:59:49 +01:00
ec896d8ca7
Cosmetic changes.
...
oo_.determinist_simulation.status is a boolean.
2015-02-13 15:58:12 +01:00
59d1c18960
Added routines missing in previous commit.
2015-02-13 10:10:41 +01:00
67be25d1a8
Rewrote homotopy for perfect foresight models.
2015-02-12 15:30:27 +01:00
f40fcb9724
Do not print anything if homotopy (perfect forsight models solver).
2015-02-12 15:30:27 +01:00
bcc68ff15c
Updated dseries submodule and dynare_config().
...
Routines used for loading data are now located in the dseries submodule.
2015-02-12 15:30:27 +01:00
a44e240153
Removed routines for loading data (dynare/matlab folder).
2015-02-12 15:30:27 +01:00
a2bd920d6f
Workaround for #755
...
(cherry picked from commit 9a3cca2a55d57c3148fbcfd892a67a1d9731e9b1)
Conflicts:
matlab/load_csv_file_data.m
2015-02-12 15:30:27 +01:00
93966f2087
Set verbosity to 0 if homotopy is used to solve the perfect foresight model.
2015-02-12 15:30:27 +01:00
751e518811
Removed global variables in perfect foresight routines.
2015-02-12 15:30:27 +01:00
892e6818ed
Put argument checks in another routine.
2015-02-12 15:30:27 +01:00
6f7fed2046
Moved sim1* routines in subfolder.
2015-02-12 15:30:26 +01:00
1b55975231
Moved det_cond_forecast routine in subfolder.
2015-02-12 15:30:26 +01:00
e4f4a03798
Removed useless routine.
2015-02-12 15:30:26 +01:00
df4ee4b983
Moved files related to perfect foresight models in a subfolder.
2015-02-12 15:30:26 +01:00
8320a39e61
Add check whether steady state file during estimation overwrittes parameters to be estimated.
...
This might lead to wrong results and should be checked for.
2015-02-11 12:58:36 +01:00
bdb6a2602e
preprocessor: ensure that options_.k_order_solver = 1 when order >= 3, closes #844
2015-02-10 03:39:18 +01:00
f72910ae33
fix preprocessor implimentation of filter_algorithm option to estimation. closes #843
2015-02-09 03:20:37 +01:00
8673bcf558
reporting: add titleTruncate option
2015-01-27 09:54:37 +01:00
1eabb1f6da
Change headers to reflect availability of third order approximation
2015-01-27 08:05:57 +01:00
5d9b25d2be
In dr_block, debugging information printed out should not be handled with options_.verbosity. Adding an additional optional argument to handle the debugging information
2015-01-21 13:48:25 +01:00
73e292df2a
merged recent changes to submodule dseries
2015-01-17 12:05:12 +01:00
8a52ef599b
added a local variable for maximum_lag
2015-01-15 13:55:48 +01:00
Johannes Pfeifer
Houtan Bastani
Sébastien Villemot
Marco Ratto
Stéphane Adjemian
Johannes Pfeifer
Michel Juillard
ferhat