Marco Ratto
e5854cb201
Fixes to mode_compute=5. (1) force univariate steps at every iteration: improve robustness at the expenses of a bit less efficiency; (2) fix gradient computation for special cases when the value of the parameters get stacked at the prior boundary.
2013-12-07 10:52:27 +01:00
Houtan Bastani
ea888fe646
write_latex_definitions: use M_ as global
2013-12-06 16:01:38 +01:00
Houtan Bastani
f70ebea94c
new function write_latex_definitions closes #478
2013-12-06 15:33:49 +01:00
Sébastien Villemot
c06a8df8a4
Merge pull request #553 from rattoma/master
...
Trap the case where only one parameter is estimated and only one non-con...
2013-12-06 06:16:50 -08:00
Marco Ratto
d969a8129c
Trap the case where only one parameter is estimated and only one non-constant row is available in Jacobian: (1) do not use vnorm; (2) do not use advanced option.
...
(cherry picked from commit d9856879ea4d9a7d9e954cc9392d1f0e98634cda)
2013-12-06 08:21:10 +01:00
Johannes Pfeifer
c9ebe21faf
Harmonize plotting of IRFS
...
In stoch_simul.m IRFs were only printed if they were not flat instead of checking whether the IRF is different from 0. Unit root responses were thus filtered out, e.g. C=0.2 for all t was not displayed.
2013-12-05 17:46:38 +01:00
Stéphane Adjemian (Charybdis)
deed4df569
Added @dseries/mpower method (element-by-element powers).
2013-12-03 23:57:53 +01:00
Sébastien Villemot
d9c3073fd9
Merge pull request #539 from JohannesPfeifer/stab_map_bivariate
...
Continue harmonizing figure titles in GSA
2013-12-03 06:50:51 -08:00
Johannes Pfeifer
51adbd8008
Update error message if SS-Anova Toolbox is missing
2013-12-03 11:45:29 +01:00
Johannes Pfeifer
3cb16af5e3
Continue harmonizing figure titles in GSA
2013-12-03 11:35:20 +01:00
Houtan Bastani
22949820be
bug fix: close unclosed file id, #438
2013-12-02 12:34:09 +01:00
Sébastien Villemot
d8d0f2b8cc
Revert cosmetic change of 4fa2d5b
2013-11-29 23:09:51 +01:00
Stéphane Adjemian (Scylla)
dc3bcb47bf
Cosmetic (removed commented line of code).
2013-11-29 19:06:26 +01:00
Stéphane Adjemian (Scylla)
a6a10055da
Fixed bug (problem with dates selection in @dseries::subsref).
2013-11-29 19:02:40 +01:00
Stéphane Adjemian (Scylla)
e1bff2c885
Adapted unitary tests for octave.
2013-11-29 19:02:40 +01:00
Stéphane Adjemian (Scylla)
59f1967347
Adapted unitary test for octave.
2013-11-29 19:02:40 +01:00
Stéphane Adjemian (Scylla)
67d71a628b
Fixed bug (isscalar does not assume that the input is numeric).
2013-11-29 19:02:40 +01:00
Stéphane Adjemian (Scylla)
e615f00bec
Fixed bug in unitary test (due to a change of behaviour introduced last week by Houtan).
2013-11-29 19:02:40 +01:00
Stéphane Adjemian (Scylla)
ee506c4ebe
Added new routine to test if a global variable exist in the base workspace.
2013-11-29 19:02:40 +01:00
Stéphane Adjemian (Scylla)
4f71d74512
Bug fix (wrong handle).
2013-11-29 19:02:40 +01:00
Stéphane Adjemian (Scylla)
121aa5038c
Bug fix. Input must be numeric in the first place.
2013-11-29 19:02:40 +01:00
Stéphane Adjemian (Scylla)
0419aa00e2
Cosmetic change (use isdates routine).
2013-11-29 19:02:40 +01:00
Stéphane Adjemian (Scylla)
8eb45edb92
Fix for Octave (need to add extension of the m script).
2013-11-29 19:02:40 +01:00
Sébastien Villemot
c1c656f4d9
Fix config of warnings in @dseries/cumsum tests
2013-11-29 17:16:17 +01:00
Stéphane Adjemian (Scylla)
45c117fbff
Fixed bug in Matlab unitary tests. Wrong dimension for returned infor variable (missing column for execution time).
2013-11-29 11:46:07 +01:00
Stéphane Adjemian (Scylla)
6e8bd16cf6
Do not define LOG if Matlab unitary tests are run from Octave.
2013-11-29 11:46:07 +01:00
Johannes Pfeifer
ec9c32d2b7
Correct name of output folder in writing TeX-Code
2013-11-29 10:25:25 +01:00
Stéphane Adjemian (Scylla)
01f2dd2ab1
Changed the paths to Dynare's prior and post processing hooks.
2013-11-28 15:21:52 +01:00
Stéphane Adjemian (Scylla)
1ff556ada2
Give more informations if the mod file is not found in the current directory.
2013-11-28 14:49:42 +01:00
Stéphane Adjemian (Scylla)
baa596f9be
Give more details if the mod file is not found in the current directory.
2013-11-28 11:07:14 +01:00
Stéphane Adjemian (Scylla)
4fa2d5b8e9
Added comments and cosmetics.
2013-11-28 10:57:54 +01:00
Stéphane Adjemian (Scylla)
be4a53e4fa
Same as commit #ee91c7654ff753728850e68366c60fe1be33ebb3 for estimation.
2013-11-28 10:54:15 +01:00
Stéphane Adjemian (Scylla)
ee91c7654f
Return a human readable message error if order=0 (stoch_simul).
2013-11-28 10:51:20 +01:00
Stéphane Adjemian (Scylla)
edb4398e28
Merge branch 'master' into fix-mcmc-algorithm
2013-11-27 17:57:36 +01:00
Stéphane Adjemian (Scylla)
130125c5da
Changed the error message issued when the initial conditions for the estimated parameters are outside the bounds (the error message depends on the value of options_.prior_trunc).
2013-11-27 12:27:51 +01:00
Stéphane Adjemian (Scylla)
fb487bc87e
internals --display-mh-history displays the overall acceptance ratios (ie using all the mcmc draws saved on disk) instead of the current acceptance ratios (ie using the mcmc draws generated by the last estimation command). Graphical and text waitbars display the current acceptance ratios. The current acceptance ratios are saved in the mh-history files.
2013-11-27 11:56:38 +01:00
Stéphane Adjemian (Scylla)
c92c5c7fb9
Cosmetic change. Renamed "acceptation rates" as "acceptance ratios".
2013-11-27 11:34:49 +01:00
Sébastien Villemot
4e4fc2ffdb
Use load_xls_file_data in read_variables (for consistency with CSV)
2013-11-26 18:19:48 +01:00
Sébastien Villemot
36ec2561bf
load_xls_file_data: under Octave, fail gracefully if io is not present
2013-11-26 18:19:48 +01:00
Sébastien Villemot
93416a93da
Improvements in load_csv_file_data
...
* autodetect whether there are variable names and dates in the file
* optimize under Octave, using csv2cell and parcecell from io package
2013-11-26 18:19:48 +01:00
Houtan Bastani
eb75397e0f
fix try / catch statement in tests for octave
2013-11-26 17:35:27 +01:00
Houtan Bastani
0d4d085447
fix try / catch statement for octave
2013-11-26 17:22:01 +01:00
Houtan Bastani
9454862142
internals: replace % for both matlab and octave with %%
2013-11-26 16:55:04 +01:00
Stéphane Adjemian (Scylla)
b8c6e4bb01
Merge branch 'master' into fix-mcmc-algorithm
2013-11-26 16:53:27 +01:00
Houtan Bastani
89268f4f6f
internals test: fix printing for octave
2013-11-26 15:29:23 +01:00
Stéphane Adjemian (Scylla)
10150a081d
Bug fix (internals command with --load-mh-history option).
2013-11-26 14:03:11 +01:00
Stéphane Adjemian (Scylla)
bd630d7210
Removed MC_record field from oo_ structure and the output argument from MCMC routines.
...
Details about the MCMC can be loaded in the workspace with the following command:
>> internals --load-mh-history <NAME_OF_THE_MOD_FILE>
under the name mcmc_informations, or printed in the command window, using the following
command:
>> internals --display-mh-history <NAME_OF_THE_MOD_FILE>
2013-11-26 12:56:41 +01:00
Stéphane Adjemian (Scylla)
80a853944a
Also display the initial value of the posterior kernel, when internals --display-mh-history command is used.
2013-11-26 12:37:14 +01:00
Stéphane Adjemian (Scylla)
b8a02d4484
Renamed record.InitialLogLiK as record.InitialLogPost.
2013-11-26 12:16:31 +01:00
Stéphane Adjemian (Scylla)
f2f072bce1
Added options to the internals command to load mh-history files or display informations about MCMC settings and state.
...
For instance, assuming that fs2000a.mod is in the current Matlab/Octave's directory and that a metropolis has been run,
the following command:
>> internals --display-mh-history fs2000a
produces the following output in the Matlab/Octave's command window:
================================
MCMC set-up for fs2000a mod file
================================
MCMC chain number 1:
--------------------
o Number of MCMC files is 1
o Number of draws is 6000
o Acceptance ratio is 35.28%
o Last value of the posterior kernel is: 1221.81656
o State of the chain:
|| Initial | Current
++++++++++++++++++++++++
e_a || 0.01570 | 0.01608
e_m || 0.00528 | 0.00512
alp || 0.35026 | 0.36909
bet || 0.99300 | 0.99308
gam || 0.00086 | 0.00038
mst || 1.00088 | 1.00147
rho || 0.65926 | 0.71132
psi || 0.66428 | 0.60611
del || 0.00977 | 0.00198
MCMC chain number 2:
--------------------
o Number of MCMC files is 1
o Number of draws is 6000
o Acceptance ratio is 33.38%
o Last value of the posterior kernel is: 1220.39229
o State of the chain:
|| Initial | Current
+++++++++++++++++++++++++
e_a || 0.01664 | 0.01576
e_m || 0.00489 | 0.00504
alp || 0.39701 | 0.36630
bet || 0.99305 | 0.99219
gam || 0.00067 | -0.00025
mst || 1.00186 | 1.00218
rho || 0.70060 | 0.63847
psi || 0.65623 | 0.74668
del || 0.00666 | 0.00842
while the command:
>> internals --load-mh-history
loads the content of the record structure (saved in the last mh-history file) in Matlab/Octave's workspace
under the name mcmc_informations.
2013-11-26 11:28:40 +01:00