Stéphane Adjemian (Lupi)
46a149576b
Merge branch 'pr#1241'
2016-07-01 16:30:49 +02:00
Stéphane Adjemian (Hermes)
2435040929
Changed default for nonlinear filters.
2016-07-01 11:47:21 +02:00
Stéphane Adjemian (Hermes)
b6ecf5fdd0
Updated particles submodule (fixes for apf and cpf).
2016-07-01 11:41:39 +02:00
Johannes Pfeifer
2906cbbddd
Add check whether toolbox is installed
2016-06-28 09:39:32 +02:00
Houtan Bastani
848d81eb84
Merge pull request #1238 from rattoma/gsa
...
Bug fixes for identification with ML
2016-06-27 23:45:05 +02:00
Stéphane Adjemian (Hermes)
fcaf457f35
Cosmetic change (code factorization).
2016-06-24 17:35:56 +02:00
Johannes Pfeifer
420e80e538
discretionary_policy: Call steady_state-file to make sure that parameter dependence is correctly handled
...
Prevents cryptic crashes due to NaN in derivatives of objective function
2016-06-23 20:50:17 +02:00
Johannes Pfeifer
317f958184
Ramsey: Test whether number of instruments matches number of missing equation
...
Otherwise cryptic crashes can result
2016-06-23 20:48:01 +02:00
Marco Ratto
be758613f4
fixed bugs that prevented the computation of the asymptotic Hessian with ML.
2016-06-19 11:09:59 +02:00
Marco Ratto
25120e54f5
added check for singular values whose eigenvectors contains only ONE column
2016-06-19 11:08:37 +02:00
Stéphane Adjemian (Hermes)
25243712ba
Fixed bug on non linux platforms + Cosmetic changes.
...
The paths to mex files was wrong.
2016-06-17 22:02:34 +02:00
Johannes Pfeifer
e214ec2976
Allow for parameter_set=mle_mode in conditional_forecast
2016-06-17 00:21:08 +02:00
Johannes Pfeifer
1c420e6056
Allow for parameter_set=mle_mode in shock_decomposition
...
Required after writing results into oo_.mle_mode for ML estimation
2016-06-17 00:21:08 +02:00
Stéphane Adjemian (Hermes)
1dea10d478
Updated dynare_config wrt #c7e848948156fe8febca22276a77c9586c565d87
2016-06-17 00:16:06 +02:00
Stéphane Adjemian (Hermes)
af9b069830
Moved isoctave routine in base directory.
2016-06-17 00:04:00 +02:00
Stéphane Adjemian (Hermes)
d80f6d9a99
Minimize the number of calls to addpath.
...
These calls have a prohibitive cost...
2016-06-16 23:55:07 +02:00
Stéphane Adjemian
3ae44efbea
Merge pull request #1218 from rattoma/gsa
...
Allow Screening with neighborhood_width sampling.
2016-06-16 22:18:39 +02:00
Stéphane Adjemian (Lupi)
33d35bc3c5
Added quantile routine in missing.
...
This is missing if the stats toolbox is not available. Integration test
TeX/fs2000_corr_ME.mod was failing because of its absence.
2016-06-16 14:33:57 +02:00
Johannes Pfeifer
f86b0a5ad6
Make dyn_forecast.m account for trend_coeffs saving coefficients for all variables
2016-06-16 12:05:38 +02:00
Johannes Pfeifer
3b09ab5424
Save Smoother.Trend and Smoother.Constant after MCMC
2016-06-16 12:05:38 +02:00
Johannes Pfeifer
1c84ccdc34
Save trend and constant for all variables in smoother
2016-06-16 12:05:38 +02:00
Johannes Pfeifer
3efe283bf0
Make oo_.Smoother.TrendCoeff have size of endogenous variables
...
Sets coefficients with unspecified trends to 0, allowing to later use the trend_var command for setting trends
2016-06-16 12:05:38 +02:00
Johannes Pfeifer
5d6d1336ef
Enable filter_covariance option in posterior sampling
2016-06-16 12:05:38 +02:00
Stéphane Adjemian (Hermes)
65ae97d37d
Fixed bug.
...
For uniform priors use the prior mean when computing the prior
moments (prior mode is not defined).
2016-06-16 11:59:47 +02:00
Stéphane Adjemian (Hermes)
789f21257e
Avoid side effects on globals (prior CLI and write_latex_prior_table commands).
2016-06-16 11:57:16 +02:00
Stéphane Adjemian (Hermes)
2c074d7cd2
Check that a prior is defined in prior CLI command
...
Behaviour consistent with the logic introduced in commit
64d01a806e
.
2016-06-16 11:16:23 +02:00
Stéphane Adjemian (Hermes)
a91e4dec75
Bug fixes.
...
Introduced when we increased the size of the mat files saved on
disk (see commit #e2942662479f822cd4ca1251158f0b5037c3a446).
2016-06-16 11:15:22 +02:00
Stéphane Adjemian (Hermes)
64d01a806e
Fixed behaviour of write_latex_prior_table command
2016-06-16 10:38:40 +02:00
Johannes Pfeifer
e94a698a9b
Implement write_latex_prior_table.m
...
Fixes and replaces write_table_prior.m and uses consistent name. Also includes documentation
2016-06-15 16:47:39 +02:00
Stéphane Adjemian (Lupi)
fcb738616c
Fixed bug in extened_path_mc.
...
Wrong calling sequence to the routine defining the shocks.
2016-06-15 16:01:08 +02:00
Stéphane Adjemian (Lupi)
1e7fc89e8d
Replaced filesep by hard coded slash.
...
See 57501f802e
.
2016-06-15 13:14:55 +02:00
Stéphane Adjemian (Lupi)
57501f802e
Revert "Cosmetic change."
...
This reverts commit 544da84e73
.
On Windows filesep returns a backslash ('\') as expected but
LaTeX, even on Windows, only understands slashes ('/') when
defining paths (for graphic or file inclusion).
2016-06-15 13:07:59 +02:00
Stéphane Adjemian (Hermes)
473d335851
Fixed bug (wrong structure names).
2016-06-15 00:35:40 +02:00
Johannes Pfeifer
029e9495ba
Fix bug in dsge_var_likelihood.m where constant was not considered for dof when Gamma function was computed
...
Does not affect the maximum of the posterior as it only affects constant term
2016-06-15 00:30:28 +02:00
Johannes Pfeifer
4486d098b5
Save dsge_var objects at the mode
2016-06-15 00:30:28 +02:00
Johannes Pfeifer
2cf882fec5
Remove comments on dynare_estimation_1.m after implementing store_smoother_results
2016-06-15 00:30:28 +02:00
Johannes Pfeifer
d08fd37986
Do not set oo_.posterior.optimization-fields when mh_posterior_mode_estimation is used
2016-06-15 00:30:28 +02:00
Johannes Pfeifer
552a9b5230
Fix bugs previously introduces in dynare_estimation_1.m related to penalty function
2016-06-15 00:30:28 +02:00
Johannes Pfeifer
2446ab02ba
Transform persistent variables of newrat.m into function arguments
2016-06-15 00:30:28 +02:00
Johannes Pfeifer
0c00e28799
Remove objective_function_penalty_base
2016-06-15 00:30:28 +02:00
Johannes Pfeifer
329b91d717
Harmonize output of objective functions
...
Closes #1149
Mirrors 1ad8df4635
2016-06-15 00:30:28 +02:00
Johannes Pfeifer
0edd8e2025
Add penalty_hessian.m and penalty_objective_function.m
2016-06-15 00:30:28 +02:00
Stéphane Adjemian (Hermes)
544da84e73
Cosmetic change.
...
Systematically use filesep instead of '/'.
2016-06-14 22:49:01 +02:00
Stéphane Adjemian (Hermes)
96128b32ab
Fixed bug (missing folder separator).
2016-06-14 21:53:14 +02:00
Stéphane Adjemian (Hermes)
2a65e7a85a
Merge branch 'master' into johannes-identification-fixes
2016-06-14 18:17:36 +02:00
Johannes Pfeifer
fff88f1dc7
display_estimation_results_table.m: Move TeXEnd inside of if-clause to prevent crashes
2016-06-14 15:50:48 +02:00
Johannes Pfeifer
b3877b3a9e
Save chain's proposal density and make load_mh_file and mh_recover load it
...
Closes #504
2016-06-14 15:09:05 +02:00
Johannes Pfeifer
faab010753
Change file ending from TeX to tex to prevent problems on Linux
...
Related to b454c28096
2016-06-14 12:13:48 +02:00
Houtan Bastani
5303d7079b
submodule update: particles
2016-06-14 12:02:15 +02:00
Johannes Pfeifer
59cbee2b56
Provide output on posterior variance decomposition
...
Closes #1219
2016-06-14 11:53:38 +02:00
Johannes Pfeifer
5ca4204806
Only set texnames in McMCDiagnostics.m if TeX is requested
2016-06-14 11:53:38 +02:00
Johannes Pfeifer
dc1c5f182c
Make gsa function correctly extract parameter (TeX) names
2016-06-14 11:53:38 +02:00
Johannes Pfeifer
6be02ef086
Add caller function info to dyn_latex_table.m
2016-06-14 11:53:38 +02:00
Johannes Pfeifer
55b2cd6f4c
Make sure underscores and backslashes are correctly handled in TeX-loaders of gsa
2016-06-14 11:53:38 +02:00
Johannes Pfeifer
7325dbe54e
Expand collect_LaTeX_Files.m to gsa redform subfolders
2016-06-14 11:52:49 +02:00
Johannes Pfeifer
a02f7cc8f3
Correct header of DsgeSmoother.m
2016-06-14 11:52:49 +02:00
Johannes Pfeifer
42f47e0f55
Correctly indent filt_mc_.m
2016-06-14 11:52:49 +02:00
Johannes Pfeifer
c3e1c01864
Enable TeX output in rmse-mapping of gsa
2016-06-14 11:52:49 +02:00
Johannes Pfeifer
cb9095977e
Allow for underscores in TeX-output of gsa
2016-06-14 11:52:49 +02:00
Johannes Pfeifer
59eec342d2
Allow for second header string in dyntable.m
2016-06-14 11:52:49 +02:00
Johannes Pfeifer
d02f8d0377
Allow for text-command in math environment by using amsmath package in collect_LaTeX_Files.m
2016-06-14 11:52:49 +02:00
Johannes Pfeifer
f76a353d65
Allow for underscores in first header element in dyn_latex_table.m
2016-06-14 11:52:49 +02:00
Johannes Pfeifer
2c4c7558f4
Add TeX output to stab_map_2.m
2016-06-14 11:52:49 +02:00
Johannes Pfeifer
11687d8c66
Expand collect_LaTeX_Files.m to gsa Output folder
2016-06-14 11:52:49 +02:00
Johannes Pfeifer
74212ab199
Delete redundant legend input argument in filt_mc_.m that creates warning message in Matlab
2016-06-14 11:52:49 +02:00
Johannes Pfeifer
ed1dbcac53
Fix setting of names and parameter names in stab_map_
2016-06-14 11:52:49 +02:00
Johannes Pfeifer
cba10b0d90
Bugfix for setting tex-names in mcf_analysis.m
2016-06-14 11:52:49 +02:00
Johannes Pfeifer
6f410f4cc1
Provide LaTeX output for model_comparison.m
2016-06-14 11:52:49 +02:00
Johannes Pfeifer
f5c08fc0e8
Create LaTeX output for Smirnov tests
2016-06-14 11:52:49 +02:00
Johannes Pfeifer
9b82c72e4f
Remove global options structure from dyntable
2016-06-14 11:52:17 +02:00
Johannes Pfeifer
849505ec62
Fix display of NaN parameters in TeX-table in ident_bruteforce.m
2016-06-14 11:52:17 +02:00
Johannes Pfeifer
f851252b25
Pass options_ to dyn_latex_table.m to make it honor noprint option
2016-06-14 11:52:17 +02:00
Johannes Pfeifer
c6f98433d2
Reset colormap to jet in plot_identification.m
...
Necessary after Matlab changed default
2016-06-14 11:52:17 +02:00
Johannes Pfeifer
73ec852433
Deal with required naming after allowing for correlations in identification in #1200
2016-06-14 11:52:17 +02:00
Johannes Pfeifer
00cdc05d58
Pass title texts to ident_bruteforce.m to make sure that previous TeX output is not overwritten by advanced option
2016-06-14 11:52:17 +02:00
Johannes Pfeifer
e294266247
Increase default MaxNumberOfBytes to 100 MB
...
Closes #330
2016-06-14 11:51:14 +02:00
Houtan Bastani
b454c28096
fix latex compilation on linux. #1207
2016-06-14 11:37:16 +02:00
Johannes Pfeifer
7ff453846a
Fix passing of compflag in dyn_mex.m
...
Fixes 83eeafa32c
2016-06-14 10:45:44 +02:00
Houtan Bastani
83eeafa32c
fix mex file compilation with MSVC. Closes #1227
2016-06-14 10:34:14 +02:00
Houtan Bastani
2bbfb8762b
fix copyright notices
2016-06-14 10:30:57 +02:00
Michel Juillard
f145a50a85
extend path: removed global options_
2016-06-09 12:02:48 +02:00
Michel Juillard
4a6e14ea95
extended path: removed option for lcppath
2016-06-09 12:02:48 +02:00
Michel Juillard
8fa29513f3
extended path: fixing bug for order > 2 and several shocks
2016-06-09 12:02:48 +02:00
Stéphane Adjemian (Hermes)
335a188dac
Fixed bug in extended path.
...
Allow innovations with zero variance (ie do not force removal of the
variables from the varexo list).
2016-06-09 10:32:25 +02:00
Marco Ratto
fbc71aa95c
Allow Screening with neighborhood_width sampling.
...
Allow calibrated values for neighborhood_width sampling to violate prior bounds.
2016-06-07 12:22:39 +02:00
Stéphane Adjemian (Charybdis)
3d16707dd9
Updated particles submodule (bug fixes).
2016-06-01 19:56:09 +02:00
Stéphane Adjemian (Charybdis)
dc1cfb4509
Updated particles submodule (bug fix).
2016-06-01 17:12:39 +02:00
Stéphane Adjemian (Charybdis)
392977d4ef
Changed default parameters for unscented transform (Particle filters).
2016-06-01 17:01:23 +02:00
Stéphane Adjemian (Charybdis)
15f83d0e18
Added new option for setting weights in CPF algorithm.
2016-06-01 16:29:53 +02:00
Stéphane Adjemian (Charybdis)
67b808207c
Added new algorithm for the estimation of nonlinear models.
2016-06-01 16:29:53 +02:00
Stéphane Adjemian (Charybdis)
272e3c96f8
Updated particles submodule.
2016-06-01 16:29:53 +02:00
Michel Juillard
e0475fd4ef
extended path: use deterministic_simulation.status instead of maxerror
2016-05-31 12:10:12 +02:00
Michel Juillard
0ce2834281
extended path: fixed boundaries for linear problem
2016-05-31 12:10:12 +02:00
Michel Juillard
5024af0099
extended path: fixed calibrated shocks
2016-05-31 12:10:12 +02:00
Michel Juillard
c14125d585
PATH: fixing bugs
2016-05-31 12:10:12 +02:00
Michel Juillard
76f008e4e4
PATH: remove calls to lcppath that doesn't work with perfect foresight models
2016-05-31 12:10:12 +02:00
Michel Juillard
d106b20b69
perfect foresight: set boundaries with PATH
2016-05-31 12:10:12 +02:00
Michel Juillard
ddb9c0657c
fixing linear complementarity problem
2016-05-31 12:10:12 +02:00
Michel Juillard
450d7b099a
fixing error status for PATH hook
2016-05-31 12:10:12 +02:00
Michel Juillard
11161063d4
removing debugging output
2016-05-31 12:10:12 +02:00
Michel Juillard
c49332df43
add support for PATH for steady state
2016-05-31 12:10:12 +02:00
Michel Juillard
c159dff19c
adding hook for PATH solver
2016-05-31 12:10:12 +02:00
Michel Juillard
b8132eb556
extended path: adding support for calibrated shock.
...
This commit is not satisfactory because it uses M_.det_shocks
for unexpected shocks. We need to improve on it soon.
2016-05-31 12:10:12 +02:00
Johannes Pfeifer
96098a17d2
Cosmetic fix to set_all_parameters.m
2016-05-30 13:31:22 +02:00
Johannes Pfeifer
118b3ccb74
catstruct.m: remove call to nargchk, which will be removed in future Matlab versions
2016-05-28 19:32:11 +02:00
Stéphane Adjemian (Charybdis)
dfbad0404d
Fixed crash in sim1_linear when periods=1.
...
Closes #1176 .
2016-05-26 22:19:36 +02:00
Stéphane Adjemian (Charybdis)
fb4dfc6da6
Fixed typo.
2016-05-24 22:32:02 +02:00
Michel Juillard
cd55ed749c
extended_path: fixing inititalization for PATH
...
(cherry picked from commit 7d29e917f77e7e20211e5be01544d86c64af8c65)
2016-05-24 18:08:33 +02:00
Michel Juillard
28c0c92ae8
fixing typos
...
(cherry picked from commit 86995a3bd478a3dc02919424aeb77e2a550a84c5)
2016-05-24 17:57:26 +02:00
Stéphane Adjemian (Charybdis)
3fa6a96bfb
Added homotopy.
...
Fixes the issue of paths containing spurious solutions for (stochastic)
perfect foresight models.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian (Charybdis)
cfc6957625
Cosmetic changes.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian (Charybdis)
8b5b7921f6
Improved solver for (S)EP.
...
Use previous solution as an initial condition for the perfect foresight problem.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian (Charybdis)
80f1baad7e
Changed solver in simul_backward_nonlinear_model.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian (Charybdis)
6c7f3f1932
Added input to pass sequence of shocks.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian(Charybdis)
7d1d860a53
Fixed ep.init option.
...
Initialization of the perfect foresight solver (in extended path) with
the solution of the first order approximation of the model was broken.
If the value of options_.ep.init is "true"" (1) then the solution of the
first order approximation is used as an initial guess for the newton
lilke solver. If the value of options_.ep.init is "false" (0) the solver
is initialized with the steady state.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian(Charybdis)
dabdda0506
Fixed wrong definition of a logical variable.
...
The second outputr returned by perfect_foresight_solver_core is the
max. abs. residual, not a dummy variable indicating success or failure
of the perfect foresight solver.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian(Charybdis)
fbf2fae036
Changed option value for unscented integration nodes.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian(Charybdis)
fdbd4fa7a7
Code factorization.
...
- Added routines for initializing and setting shocks in EP.
- Added a specialized routine for doing Monte Carlo around EP.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian(Charybdis)
3bbac629ed
Cosmetic change.
...
Removed maximum_lead and maximum_lag in extended_path routines.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian(Charybdis)
d74f315bfb
Fixed bug in (Stochastic) Extended Path algorithm.
...
Innovations were not correctly passed to the non linear solver. Closes #1128 .
2016-05-24 17:28:24 +02:00
Stéphane Adjemian (Hermes)
4926b6c6ff
Added a routine the residuals of a model.
...
Takes a path for the endogenous variables and returns the residuals of
the dynamic equations. This routine is useful for solving the model with
the PEA approach.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian (Hermes)
ea1ebfa911
Added argument to simul_backward_nonlinear_model.
...
First input argument is a vector for the initial condition of the
endogenous variables. If empty, the steady state of the model is used.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian (Hermes)
e507a6d0a9
Fixed copyright header.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian (Hermes)
8958b7ef98
Cosmetic changes.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian (Hermes)
3a1f74d7a7
Fixed typo.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian (Hermes)
13ca15a278
Allow (S)EP with arbitrary sequence of innovations.
...
The third input argument of extended_path Matlab/Octave's routine is the
sequence of shocks (T*n array, where n is the number of exogenous
variables and T is the size of the sample). If the third argument is
empty, the (stochastic) extended path is run with gaussian
innovations (this corresponds to the previous behaviour).
TODO:
- Fix the compatibility with ep.replic_nbr
- Check the 'calibrated' mode.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian (Hermes)
b60bd7b36b
Removed globals from extended_path routine.
2016-05-24 17:28:24 +02:00
Stéphane Adjemian (Hermes)
ee78ad2049
Improved execution speed.
...
Removed some calls to bsxfun which is not optimal on sparse matrices. I
only removed the calls where I identified bottlenecks (with a Smets and
Wouters model), more tests are needed to decide if we have to remove other
occurences of bsxfun on sparse matrices.
2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Hermes)
be0f6230ad
Cosmetic changes.
...
- Use a switch-case block.
- Added a warning when the user tries to solve a linear(ized) model
with solve_algo different from 0 (for a linear model).
- Added an error message if the user try to solve a linearized model
with stack_solve_algo=6 (not implemented).
2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Hermes)
1815eb8e56
Code factorization.
2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Hermes)
8a163842d0
Removed oo_ from sim1_linear routine.
2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Hermes)
53747cb082
Fixed lmmcp option.
...
This option may be used in extended_path and perfect_foresight_solver commands.
2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Charybdis)
e0be60710c
Removed oo_ from sim1_purely_{backward,forward} routines.
2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Charybdis)
290ce5db68
Removed oo_ from sim1_lbj routine.
2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Charybdis)
674c9ab286
Removed oo_ from sim1 routine.
2016-05-24 17:26:06 +02:00
Marco Ratto
3e8dbcfc7e
Set mode_files in such a way that it can be handled with key-value pair.
...
We need a single mode file that contains variable xparams = [x1 x2 x3] all multiple modes, one per column.
Fixes #1199
2016-05-22 23:15:18 +02:00
Marco Ratto
d8fea3ff13
Correlations and measurement errors in identification:
...
1) Filter out measurement errors with error message that suggests to explicitly write measurement errors in model definition
2) allow identification checks with correlations, by switching to numerical derivatives
2016-05-22 23:10:01 +02:00
Johannes Pfeifer
389f43019d
Put fields of OSR in M_ into workspace to restore compatibility with 4.4.3
2016-05-22 11:41:54 +02:00
Johannes Pfeifer
d10f8b42d7
Fix bug in convert_dyn_45_to_44.m if bayestopt_ is empty
2016-05-22 11:41:54 +02:00
Johannes Pfeifer
03563e012b
Allow for pass OSR parameter bounds to optimizer
2016-05-22 11:41:54 +02:00
Johannes Pfeifer
9774d0544d
Initialize fields of M_.osr to empty in global_initialization.m
2016-05-22 11:41:54 +02:00
Johannes Pfeifer
676182f550
Cosmetic fix to dynare_minimize_objective.m
2016-05-22 11:41:54 +02:00
MichelJuillard
3df1a1275f
Merge pull request #1185 from JohannesPfeifer/bytecode_fixes
...
Bytecode fixes
2016-05-22 10:32:56 +02:00
Johannes Pfeifer
13d42e05a5
Add missing else-clause in simulated_moment_uncertainty.m
...
Fixes crash in unit tests
2016-05-22 10:03:56 +02:00
Stéphane Adjemian
0521aa4e09
Merge pull request #1188 from JohannesPfeifer/slice_marco
...
Add slice and factorization of posterior samplers
2016-05-22 01:45:14 +02:00
Johannes Pfeifer
161f674dd9
Fix two TeX loaders where apostrophe was missing
2016-05-20 12:43:52 +02:00
Johannes Pfeifer
7c97d7c662
Allow displaying 0 rows in disp_dr.m
2016-05-20 12:42:43 +02:00
Johannes Pfeifer
2baf88da47
Remove adaptive_metropolis_hastings
2016-05-19 23:16:29 +02:00
Johannes Pfeifer
89fd14f026
Filter out unit root cases that lead to crashes
2016-05-19 18:57:47 +02:00
Johannes Pfeifer
1f71331e58
Fix dynareParallelGetFiles.m call to random_walk_metropolis_hastings
2016-05-19 17:43:48 +02:00
Johannes Pfeifer
12ee5b7276
Fix copyrights and references to random_walk_metropolis_hastings headers
2016-05-19 17:15:31 +02:00
Johannes Pfeifer
781ea45777
Remove old independent_metropolis_hastings routines
2016-05-19 14:37:05 +02:00
Johannes Pfeifer
3073988ac1
Fix vector dimension for independent_metropolis_hastings
2016-05-19 14:37:05 +02:00
Johannes Pfeifer
86fcde0879
Cosmetic changes to dynare_estimation_1.m and store_smoother_results.m
2016-05-19 14:37:05 +02:00
Johannes Pfeifer
4898a66d2a
Allow passing of optim option for TaRB
2016-05-19 14:37:05 +02:00
Johannes Pfeifer
90ab263587
Fix location of TaRB options
2016-05-19 14:37:05 +02:00
Johannes Pfeifer
413a264b21
Make read_key_value_string.m compatible with sublists
2016-05-19 14:37:05 +02:00
Johannes Pfeifer
091f94221f
Delete unused TaRB_metropolis_hastings_core.m
2016-05-19 14:37:05 +02:00
Johannes Pfeifer
c7e8489481
Rename autocorr into dyn_autocorr.m as it does not mirror the behavior of the econometric toolbox function
2016-05-19 14:37:05 +02:00
Marco Ratto
1d6961cc4c
initialize mode field in slice sampler options
2016-05-19 14:37:05 +02:00
Marco Ratto
e5ab7ed992
fixed unknown options warning messages
2016-05-19 14:37:05 +02:00
Marco Ratto
088db78118
Document computation of Inefficiency factors
2016-05-19 14:37:05 +02:00
Marco Ratto
ebe34bc40f
Make display of inefficiency factors use existing table util [thanks to Johannes]
2016-05-19 14:37:05 +02:00
Marco Ratto
7b3c42c6e1
provisions for reworked posterior sampling options:
...
- handle sub lists of individual samplers
- split checks in dynare_estimation_init.m and before running posterior_sampler.m [invhess checks]
- posterior sampler options checks moved from initial_estimation_checks.m to check_posterior_sampler_options.m
- added use_mh_covariance_matrix to imh and rwmh
- slice re-sets mode_compute=0 cova_compute=0
- updated test function
2016-05-19 14:37:05 +02:00
Houtan Bastani
cf6e3cb14c
preprocessor: rework posterior sampler options, #1177
2016-05-19 14:37:05 +02:00
Marco Ratto
35a1c7ac17
Commented options for slice sampler, including default values and associated restrictions/default for main options_ fields
2016-05-19 14:34:09 +02:00
Marco Ratto
9c62021d3c
Move TaRB to posterior_sampler_iteration.m [PATCH from Johannes: thanks a lot]
2016-05-19 14:34:09 +02:00
Marco Ratto
0c01565298
Allow sampler specific options via optim;
...
bug fixes on options logic
2016-05-19 14:34:09 +02:00
Marco Ratto
22f51d1e75
moved autocorr to missing matlab econometrics toolbox [it is not in stats toolbox]
2016-05-19 14:34:09 +02:00
Marco Ratto
4445f17efa
Manual merge with commit history of metropolis_hastings_initialization.m and random_walk_metropolis_hastings_core.m
2016-05-19 14:34:09 +02:00
Marco Ratto
357e3c2ac7
Make posterior_sampler_initialization.m compatible with new form of prior_draw.m
2016-05-19 14:34:09 +02:00
Marco Ratto
8e751be5ee
bug fix for non integer Nc
2016-05-19 14:34:09 +02:00
Marco Ratto
e69c5f6864
Bug fix for odd Nc
2016-05-19 14:34:09 +02:00
Marco Ratto
8bd963de64
Fill in posterior_mode with info from posterior samples.
2016-05-19 14:34:09 +02:00
Marco Ratto
b830805cb6
Trap empty d [e.g. from slice sampler]
2016-05-19 14:34:09 +02:00
Marco Ratto
a96dd5b964
for slice, always initialize empty invhess unless a new chain is initialized from a mode file
2016-05-19 14:34:09 +02:00
Marco Ratto
9089fc340c
Initialize posterior_sampler_options.
2016-05-19 14:34:09 +02:00
Marco Ratto
46908d20c6
Provisions for new posterior sampler routines that factorize the individual iteration of individual samplers.
...
I contains a quick fix to adaptive_posterior_sampler.m, which cannot be embedded in the new structure.
2016-05-19 14:34:09 +02:00
Marco Ratto
c30b6f20a6
Added routines that factor out posterior samplers into the iteration routine posterior_sampler_iteration.m.
...
It generalizes the random_walk_metropolis_hastings.m structure but adds two new items:
- check_posterior_sampler_options.m which is a wrapper for method specific options (stored in options_.posterior_options)
- posterior_sampler_iteration.m which is a wrapper of the individual iteration of any posterior sampler.
It already embeds, with parallelization implementation:
- random_walk metropolis hastings
- independent metropolis hastings
- slice sampler
It also contains a quick fix for TARB, although the latter should be embedded in posterior_sampler_iteration.m as well.
Any new posterior sampler can be simply added to posterior_sampler_iteration.m, with wrapper to specific iterations in check_posterior_sampler_options.m
2016-05-19 14:34:09 +02:00
Marco Ratto
b1b901fc61
Added SLICE sampler core routines
2016-05-19 14:34:09 +02:00
Marco Ratto
951dfe8ff4
When mh_posterior_mode_estimation, trap cases where the prior mode is at the boundary.
2016-05-19 14:34:09 +02:00
Marco Ratto
a3ab6de6c8
Added function to compute of inefficiency factors, used in McMCDiagnostics.m
...
Requires missing autocorr (available in some MATLAB toolboxes).
Requires factor out CutSample in dynare_estimation_1.m
2016-05-19 14:34:09 +02:00
Johannes Pfeifer
0e2e1376c9
Fix bug in simulated moments
...
Closes #1105
2016-05-18 22:51:20 +02:00
Houtan Bastani
58eb60c95d
fix copyright dates, headers
2016-05-18 15:06:49 +02:00
Houtan Bastani
6a0de7a724
Merge pull request #1189 from rattoma/gsa
...
gsa and identification fixes
2016-05-18 15:01:19 +02:00
Johannes Pfeifer
c05c66cbaa
Filter out error code 6 in dsge_var_likelihood.m
...
Crashes otherwise
2016-05-17 20:56:25 +02:00
Marco Ratto
42170c6c13
Fixes around models with unit roots.
2016-05-17 11:47:07 +02:00
Marco Ratto
c9bc685049
Fix determination of full rank subspace of simulated moment uncertainty
2016-05-17 09:38:06 +02:00
Marco Ratto
f0b346b397
added waitbar, to monitor computations [especially useful for large models]
2016-05-17 08:41:55 +02:00
Marco Ratto
7c8878c071
Allow neighborhood_width sampling in Morris + bug fix.
2016-05-17 08:41:16 +02:00
Marco Ratto
28aeb9f308
bug fix
2016-05-17 08:40:32 +02:00
Houtan Bastani
5723e7ea58
build system: remove support for 32-bit Matlab on OS X
2016-05-13 10:53:55 +02:00
Houtan Bastani
89667ea8da
remove unused file. closes #1172
2016-05-11 15:12:27 +02:00
Michel Juillard
31f8d61586
make estimation honor steady state nocheck option
2016-05-10 15:54:11 +02:00
Houtan Bastani
ccb197e74f
fix copyright dates
2016-05-09 12:42:58 +02:00
Michel Juillard
5e5157e375
loglinear: bug correction
2016-05-06 10:12:12 +02:00
Michel Juillard
67ea280773
adding missing semi-colon
2016-05-06 10:11:53 +02:00
Michel Juillard
ef232ac436
fixed log_variable function
2016-05-05 20:53:22 +02:00
Michel Juillard
3c7c2f46a4
more fixes for loglinear with lead/lagged exogenous variables
2016-05-05 20:41:02 +02:00
Michel Juillard
7e13fe5ef7
fix loglinear option with leaded or lagged exogenous variables
2016-05-05 20:41:02 +02:00