Commit Graph

7835 Commits (7b1663281bc101b86d0c48bc1563208693ab08fe)

Author SHA1 Message Date
Johannes Pfeifer 7b1663281b Transform objective function in osr to full matrix
If the objective function only has one element, the scalar product of a sparse matrix and a full matrix is sparse (in contrast to two or more elements). This crashes the osr code subsequently.
2014-11-03 15:29:21 +01:00
Johannes Pfeifer 4aaebfbe46 Add unit test to check whether objective function matches and whether covariances actually work 2014-11-02 13:15:55 +01:00
Johannes Pfeifer ecbe428bd7 Fix computation of objective function in OSR if covariances are specified
Due to the preprocessor adding entries for all variable combination into i_var, variables contained in covariances obtained a wrong weighting. By only selecting the unique entries, this cannot happen anymore. Moreover, the computation intensity of the objective does not increase quadratically in the objective anymore.
2014-11-02 12:49:11 +01:00
Stéphane Adjemian (Charybdis) 9300a061aa Fixed bug (appearing only with Matlab and legacy data interface) if the m data file is in the current directory.
If the m data file is in the current directory, the first output of
fileparts(options_.datafile) is an empty string. In Matlab the command
cd('') fails, but Octave does not fail in this case (and does nothing).

Added a test on the first output of fileparts, if this output is empty
we do not try to change the current directory.
2014-10-23 11:08:26 +02:00
Stéphane Adjemian (Karaba) 7ab86d31ec Dataset factorization. Removed tests/kalman/lyapunov/fsdat_simul.m and use tests/fs2000/fsdat_simul.m instead. 2014-10-21 16:42:52 +02:00
Stéphane Adjemian (Karaba) 5ab08ed8d3 Added missing semi colon. 2014-10-21 16:37:57 +02:00
Stéphane Adjemian (Karaba) ca28a2b778 Fixed load_m_file_data_legacy.m.
An m file data was not loading if outside the current directory.
2014-10-21 16:05:34 +02:00
Stéphane Adjemian (Karaba) 63d3157267 Do not force the user to provide a clean m data file when using the legacy data interface.
Note that  all the variables  listed in varobs  need to have  the same
number of observations.
2014-10-20 17:28:35 +02:00
Stéphane Adjemian (Karaba) a06083cb12 Fixed test (new data interface). 2014-10-20 16:31:40 +02:00
Stéphane Adjemian (Karaba) f48566aeae Fixed prior bounds (according to the doc in master branch).
* Second  and  third  positional  arguments  after the  name  of  the
   estimated  parameter   in  the  estimated_params   block  are  only
   considered in the optimization stage (not in the MCMC)

 * Do not  store bounds  in bayestopt_, because  bounds do  not always
   reflect restrictions implied by prior shapes.

 * prior_bounds routine  returns a structure  (with fields lb  and ub)
   instead of a matrix.
2014-10-20 16:18:54 +02:00
Stéphane Adjemian (Karaba) fb14d24036 Do not truncate the plotted prior densities with the truncation
parameters  used   for  optimization  (second   and  third  positional
arguments  after   the  name  of   the  estimated  parameter   in  the
estimated_params block).
2014-10-16 15:27:40 +02:00
Stéphane Adjemian (Karaba) 2d61c39276 Cosmetic changes. 2014-10-16 15:25:51 +02:00
Stéphane Adjemian (Karaba) d423ae63a8 Set default value for last input argument of lyapunov_symm routine (debug flag). 2014-10-15 09:02:42 +02:00
Michel Juillard c8ac7007f0 fixing bug in test case of standalone c++ 2014-10-14 10:25:24 +02:00
Stéphane Adjemian (Charybdis) 8244862bd1 Removed "spurious" characters. 2014-10-13 23:01:40 +02:00
Stéphane Adjemian (Charybdis) 9b3337c94d Oups. I forgot to remove a dependency to a binary in tests/Makefile.am. 2014-10-13 22:21:23 +02:00
Stéphane Adjemian (Karaba) 935f74883d Added unit tests for lyapunov_symm routine. 2014-10-13 17:42:53 +02:00
Stéphane Adjemian (Karaba) f99fcfa27f Removed "Do pure Lyapunov tests" section.
This should be done in the lyapunov_symm routine.
2014-10-13 17:42:53 +02:00
Stéphane Adjemian (Karaba) 97b63105a0 Add a parameter to the lyapunov_symm routine (debug mode). 2014-10-13 17:42:53 +02:00
Stéphane Adjemian (Karaba) a9b5fb4263 Fixed test introduced in commit and removed binary file.
Test was introduced in b11d14498aadac17c08b08dd80b6daf7dec9e340.

The binary  file was a mat  file containing a posterior  mode, and the
second estimation must be run iff the control toolbox is available.
2014-10-13 17:42:53 +02:00
Stéphane Adjemian (Karaba) 55808060cf Removed useless input (complete commit 438a671c3873e5c4e1bb02a54080e930fbc11886). 2014-10-13 17:42:52 +02:00
Johannes Pfeifer 7605820dff Add unit test for Lyapunov solver
- tests whether they run within a mod-file
- tests whether they yield the same result for a simple state-space model
2014-10-13 17:42:52 +02:00
Johannes Pfeifer b90f3deed2 Bugfixes and improvements related to method 3 of lyapunov_symm.m
- the function was called with the wrong input argument for this case (Q instead of R*Q*R'), crashing with incompatible matrix dimensions
- the persistent variable X was not updated when the size of it changed, leading to crashes when estimation is followed by calls to DsgeSmoother.m where the state-space is different
- Also eliminates the printed output in lyapunov_symm.m that could not be turned off and clutters estimation
2014-10-13 17:42:52 +02:00
Houtan Bastani 466a6d23a8 preprocessor: add field to store the number of multiplier equations added to ramsey problem. closes #728 2014-10-13 17:25:08 +02:00
Houtan Bastani 3e8acef4cc preprocessor: add field that contains the total number of equations, #728 2014-10-13 17:17:50 +02:00
Houtan Bastani da7d721063 preprocessor: always output the original number of equations 2014-10-13 17:16:41 +02:00
Houtan Bastani 70d00bf2ee remove repetitive none statement introduced in a6009908e4 2014-10-13 16:02:37 +02:00
Stéphane Adjemian 573d276f1a Merge pull request #739 from JohannesPfeifer/Cosmetic_changes
Cosmetic changes, documentation, and small bugfix
2014-10-13 13:39:19 +02:00
Johannes Pfeifer 96df3b0043 Cosmetic changes to error messages
(cherry picked from commit ffa424d9d88dcd534272b154a3d9848df861914b)
2014-10-13 13:05:46 +02:00
Stéphane Adjemian (Karaba) bc1243f501 Removed config for matlab files in .dir-locals.el (for emacs users). 2014-10-13 12:52:51 +02:00
Stéphane Adjemian 78621c03ee Merge pull request #742 from JohannesPfeifer/second_order_hessian_problems
Deal with NaN and Inf in Hessian of dynamic model
2014-10-13 12:49:38 +02:00
Stéphane Adjemian 42bcd22aa2 Merge pull request #756 from JohannesPfeifer/CutSample
Fix bug still present in CutSample.m
2014-10-13 12:37:51 +02:00
Johannes Pfeifer de5875a75b Fix bug still present in CutSample.m
#733 was faulty because it did not account for changes in the first line when more than 1 MH file was present. Now, FirstDraw really denotes the first draw (as opposed to the last draw as before). The subsequent computation of FirstMhFile is then also correct.
2014-10-12 11:33:55 +02:00
Stéphane Adjemian 6c112a2701 Merge pull request #743 from JohannesPfeifer/debug_info_estimation
Add check for NaN parameters to initial_estimation_checks.m
2014-10-10 22:30:12 +02:00
Stéphane Adjemian 06fbf2b2c5 Merge pull request #746 from JohannesPfeifer/rplot
Filter out empty oo_.endo_simul in rplot.m
2014-10-10 22:25:07 +02:00
Stéphane Adjemian (Charybdis) df1e734106 Merge branch 'steady_debug' 2014-10-10 21:47:24 +02:00
Stéphane Adjemian (Charybdis) f4922a8a27 Cosmetic changes. 2014-10-10 21:46:50 +02:00
Stéphane Adjemian 9e3edd6673 Merge pull request #748 from JohannesPfeifer/maxit
Change maxit default value and correct manual
2014-10-10 19:03:56 +02:00
Stéphane Adjemian ee4586d42e Merge pull request #752 from JohannesPfeifer/doc_initval
Expand documentation on varexo in initval
2014-10-10 19:02:27 +02:00
Stéphane Adjemian 37d4cf7e35 Merge pull request #753 from JohannesPfeifer/MCMC_diag
Improve tractability of McMCDiagnostics_core.m by adding header and comm...
2014-10-10 18:46:30 +02:00
Houtan Bastani f4f7a02fd7 Provisions for MATLAB 8.4 (R2014b) 2014-10-10 10:41:04 +02:00
Johannes Pfeifer 1d68ed8da6 Improve tractability of McMCDiagnostics_core.m by adding header and comments. Also uses more intuitive variable names 2014-10-09 10:57:28 +02:00
Johannes Pfeifer c6438122a1 Expand documentation on varexo in initval 2014-10-06 09:03:35 +02:00
ferhat eecf154cc1 Correction of a bug when "block" option is used without "bytecode" option in the "model" statement: for consecutive blocks that are evaluated, only the first one is in fact evaluated but not the following ones 2014-10-04 12:19:52 +02:00
Houtan Bastani f4e40dcf83 more explicit warning when dividing by zero in conversion to static model and when reading model #730 2014-10-03 17:04:07 +02:00
Johannes Pfeifer 3187c70b97 Add debugging info on initial values for steady state computation 2014-09-30 15:05:12 +02:00
Johannes Pfeifer 1ce39ac311 Document new default values for maxit and update manual to reflect that maxit options of deterministic solvers and steady state solvers are different now 2014-09-30 11:48:12 +02:00
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
Michel Juillard 6bbfa2fa6d removed 1e6 in expression forcing a non-singular Jacobian 2014-09-30 08:13:31 +02:00
Michel Juillard 8f47c731dc fixing bug introduced in commit 4185e8af14 2014-09-30 07:58:51 +02:00