69b385ccd4
Bugfix in PosteriorIRFs
...
Fixed ancient copy and paste mistake from stoch_simul for relative IRFs
in Posterior IRFs
2013-02-28 12:00:40 -05:00
46a59423f2
Revert "Remove unused options_.verbosity"
...
This reverts commit cacb0470f8#295 .
2013-02-28 12:20:21 +01:00
0219301e23
reporting (WIP)
2013-02-27 13:18:53 +01:00
a64ec3b156
Reorganization of warnings
...
- at the end of the computation, don't display all preprocessor warnings but
only their number
- also display if there has been a MATLAB/Octave warning (counting their
number does not seem feasible). Closes #181
- add new "nowarn" option to disable all warnings. Closes #301
2013-02-26 16:50:05 +01:00
cacb0470f8
Remove unused options_.verbosity
2013-02-26 15:14:33 +01:00
bbd95b3a60
adding Ed Herbst fast implementation of the Kalman filter and test
...
cases with timing. Still needs preprocessor interface (option) and documentation.
2013-02-21 17:47:16 +01:00
78f8f04598
reporting (WIP)
2013-02-18 18:16:44 +01:00
4fa141762d
fix copyright dates
2013-02-15 11:58:30 +01:00
3c69ee04eb
new file that returns only the index for resampling particles
2013-02-15 10:35:32 +01:00
22a29ab068
modification of the resampling procedure: only one call for all methods
2013-02-15 10:35:32 +01:00
155f6f3bf7
modification in the call of resampling: only one call now
2013-02-15 10:35:32 +01:00
9df3857966
Huge modification of the file: uses now the PCA to orthogonalize the state variables to resample
...
in order to use independent smooth resampling; modification of the input of the procedure since
now no partition is required
2013-02-15 10:32:58 +01:00
d440767b03
minor modification in the call of the smooth resampling procedure
2013-02-15 10:32:58 +01:00
fefb3528c4
modification to make it comformable with the changes in resample.m
2013-02-15 10:32:58 +01:00
51c5363850
modification of the output to make it conformable with the modification in resample.m
2013-02-15 10:32:58 +01:00
920d2af8d3
modification of the resampling procedure: only one call for all
2013-02-15 10:32:58 +01:00
1d1137bf4f
reporting (WIP)
2013-02-14 18:14:44 +01:00
fe5f61db61
reporting (WIP)
2013-02-12 14:53:36 +01:00
73218bd88e
bug fix to properly trap different reasons why unique solution is not obtained
2013-02-08 14:31:49 +01:00
012c1743dd
bug fixes for when run_smoother==0
2013-02-08 14:30:40 +01:00
931b22cb26
option kalman_algo was not properly managed by gsa toolbox
2013-02-08 14:29:52 +01:00
fac7319284
@dynDate.display(): simplify the code by reusing format()
2013-02-07 15:28:43 +01:00
cfa28a58d3
Add display method for dynDate and dynDates classes
2013-02-05 14:49:30 +01:00
3a599aee84
Add subsasgn methods to dynDate and dynDates
...
The rationale is that, in the absence of a subsasgn method, MATLAB and Octave
behave differently. The former refuses any data member update, while the latter
allows all. Providing an implementation is therefore required to have uniform
behavior.
2013-02-05 14:32:08 +01:00
53f7ab33f3
Remove obsolete dynTime class
2013-02-05 14:30:49 +01:00
597b850890
Fix typo
2013-02-05 12:49:41 +01:00
68dff500ca
bug fix: typo
2013-02-05 11:18:43 +01:00
f0ef6c930a
Removed call to gaussian_filter_bank routine in sequantial_importance_particle_filter.
2013-02-05 10:37:11 +01:00
aa9bbf8f06
Add missing semicolons
2013-01-30 16:58:15 +01:00
bb9c9c566d
Bug fix when there are no shocks estimated.
2013-01-23 11:53:07 +01:00
af6bc73695
Stores the status of a deterministic simulation using bytecode in oo_.deterministic_simulation.status
2013-01-18 17:05:10 +01:00
fb6762a972
Considers a pathological case where a static model is dynamically simulated
2013-01-18 17:03:12 +01:00
47022b12c3
Considers a pathological case where the dynamic model is simulated on one period
2013-01-18 17:02:15 +01:00
fb127d010e
Correction of a bug in controlled exogenous variables indexation
2013-01-18 17:00:04 +01:00
eed3651d6e
check only returns eigenvalues, not the whole oo_
2013-01-15 16:45:15 +01:00
14ed3d94a9
Ensure that running check without stoch_simul still sets oo_.dr.eigval
...
This is documented in the reference manual, but was not working.
The change consists in having check.m returning oo_ as 1st output argument, and
having the preprocessor generating the corresponding code.
2013-01-15 16:14:15 +01:00
00485425c1
Fix crash in conditional_forecast
2013-01-15 12:48:56 +01:00
9334d762e1
* matlab/@dynSeries/private/horzcat2.m: added missing update of 'time' member
2013-01-13 11:01:32 +01:00
a74069c91d
* matlab/@dynDates/append.m: added the possibility to append a
...
dynDates object. Duplicate dates are permitted and not checked for.
2013-01-13 10:53:23 +01:00
842b8875f2
* matlab/@dynDates/subsref.m: adding sub_sample subsref to dynDates
2013-01-13 10:24:16 +01:00
e6541dabcc
* matlab/@dynSeries/subsref.m: removed spurious code introduced in
...
previous commit
2013-01-13 09:14:57 +01:00
4f7606f5a1
* matlab/@dynDate: started changing references to dynTime by
...
dynDates; dynDate can be initialized with a single dynDate
2013-01-12 21:39:54 +01:00
15fa7fc840
* matlab/@dynDate/colon.m: fixed bug when there is only one series
2013-01-12 21:36:01 +01:00
11e151547c
Adds conditional forecast using the extended path method
2013-01-11 18:04:46 +01:00
a4681b5092
Don't fail when all values are NaN/Inf
2013-01-07 17:03:47 +01:00
5c88338965
Fix mode_compute=3 under Octave
2013-01-07 15:40:48 +01:00
8c7e3d46ed
Document when mode_compute={1,3,7} are available, and error with an informative message if not
2013-01-07 15:40:45 +01:00
ad8b615771
Use maximum_endo_lag instead of maximum_lag
...
Closes : #292
2013-01-07 10:40:49 +01:00
323f455a38
Merge remote-tracking branch 'ratto/master'
2012-12-19 17:50:41 +01:00
9dfe2c908f
Added methods to compute yearly and quaterly growth rates and differences (dynSeries).
2012-12-19 17:13:52 +01:00
Johannes Pfeifer
Sébastien Villemot
Houtan Bastani
Michel Juillard
Frédéric Karamé
Marco Ratto
Stéphane Adjemian (Charybdis)
Ferhat Mihoubi