Houtan Bastani
f3ec84971b
preprocessor: add diffuse_filter and prior_trunc options to identification, #677
2014-11-18 11:53:34 +01:00
Marco Ratto
e3ac15458a
Properly deal with dseries in identification. This fixes #781
2014-11-18 11:34:49 +01:00
Houtan Bastani
8d96310436
Merge pull request #702 from rattoma/master
...
Use the nodecomposition option which is available in th_autocovarianc...
2014-11-17 12:16:43 +01:00
Houtan Bastani
4b338b4bcb
preprocessor: add nodecomposition option to estimation and stoch_simul, #702
2014-11-17 12:15:38 +01:00
Houtan Bastani
c3fdb7517c
Merge pull request #782 from JohannesPfeifer/perfect_foresight_message
...
Cosmetic change to perfect foresight + unit tests
2014-11-17 11:32:07 +01:00
Johannes Pfeifer
b735042f05
Fix naming of option selected_variables_only in manual
2014-11-16 21:11:06 +01:00
Johannes Pfeifer
c5a0f932c2
Change wrong Windows-style file separators to Linux ones
2014-11-16 21:11:06 +01:00
Johannes Pfeifer
c5c5caace5
Add utility to convert Dynare 4.5 results structure to 4.4
2014-11-16 21:11:06 +01:00
Johannes Pfeifer
cb4fb6aaf1
Replace eval-command in dynare_estimation.m by direct call to fields
2014-11-16 21:11:06 +01:00
Johannes Pfeifer
86cac40362
Harmonize field length for FilteredVariables for calibrated smoother
...
Analogous to 4f5e0321228c0e4aca19fcd114f26dbaa1bbbfaf
2014-11-16 21:11:06 +01:00
Johannes Pfeifer
4b1e815728
Always display value of posterior/likelihood at the mode
...
Also clarifies that minus the posterior/likelihood is displayed
2014-11-16 21:11:06 +01:00
Johannes Pfeifer
747f90d89b
Add unit tests for correct dimension of filtered, smoothed, forecasted, and updated variables and shocks
2014-11-16 21:08:18 +01:00
Johannes Pfeifer
9d25963976
Document different in FilteredVariablesKStepAhead-matrix between classical and Bayesian smoother
2014-11-16 21:08:18 +01:00
Johannes Pfeifer
68b27252d7
Fix selected_variables_only option with classical forecasts
...
dyn_forecast requires that the smoother has been run on all endogenous variables, leading to crashes. Now, if this option is encountered with classical forecasts, it is overridden. Also documents this in the manual.
2014-11-16 21:08:18 +01:00
Johannes Pfeifer
45ef721d03
Fix computation of SmoothedMeasurementErrors in Bayesian smoother
2014-11-16 21:08:18 +01:00
Johannes Pfeifer
1cbb17f929
Make sure classical filtered variables have the same length as the Bayesian ones, i.e. nobs.
2014-11-16 21:08:18 +01:00
Johannes Pfeifer
3902afa783
Improve documentation on filtered variables
2014-11-16 21:08:18 +01:00
Johannes Pfeifer
49abff9e4d
pm3.m: If no TeX names are provided, default to variable names provided
...
Old default assumes that only moments for endogenous variables considered. This leads to wrong results if e.g. shocks are considered.
2014-11-16 21:08:18 +01:00
Johannes Pfeifer
61253bcb81
Add header with info to pm3.m
2014-11-16 21:08:18 +01:00
Johannes Pfeifer
9f20aaa8e4
Cosmetic fix to wording of error message in resol.m
2014-11-16 21:08:18 +01:00
Johannes Pfeifer
b76bc6d79e
Make sure all computed moments in Bayesian estimation return column vectors
...
The HPD were the only exemption, being row vectors.
2014-11-16 21:08:18 +01:00
Johannes Pfeifer
5062f2ec7b
Fix length of stored and plotted forecasts for Bayesian estimation
...
It stored the initial condition(s) in addition to the forecasts for the specified horizon. In plots, the first point was actually not the forecast, but the initial condition. This contrasted with the behavior or ML and recursive Bayesian forecasts. Now the stored and plotted forecasts always have the first forecast horizon as the first entry/data point
2014-11-16 21:08:17 +01:00
Johannes Pfeifer
704f8650af
Properly initialize var_yf in forcst.m
2014-11-16 21:08:17 +01:00
Johannes Pfeifer
ac085f943a
Add two unit test for perfect foresight simulation with only forward-looking variables
2014-11-16 15:09:16 +01:00
Johannes Pfeifer
af9e30c658
Change display of failed solution to a warning
...
Problem can be missed too easily otherwise
2014-11-16 14:35:58 +01:00
Marco Ratto
a871441896
1) Forced nodisplay to avoid proliferation of plots;
...
2) added waitbar;
2014-11-14 18:00:41 +01:00
Marco Ratto
21bf34e609
1) Added new improved plotting utility for Monte Carlo filtering tests: scatter_mcf.m;
...
2) fixed graphical and printed output for prior stab_map_.m;
3) use corrcoef for pvalue and change default pvalue_corr, alpha2_redform, alpha2_rmse to 1.e-5
2014-11-14 17:59:08 +01:00
Marco Ratto
6a4f79684c
Merge remote-tracking branch 'remotes/personal/gsa' into gsa
2014-11-14 17:47:58 +01:00
Houtan Bastani
3f68450d51
osr: add option opt_algo. #779
2014-11-14 17:35:03 +01:00
Marco Ratto
06572f26a4
Use the nodecomposition option which is available in th_autocovariances.m: for large models with the HP filter option, this may spare a lot of time, if only the unconditional moments are needed on output;
2014-11-14 17:28:17 +01:00
Stéphane Adjemian (Telemachus)
aae7c23fe2
Added unitary test for hessian routine.
2014-11-14 16:33:34 +01:00
Johannes Pfeifer
ff3c8e94e1
Clean up hessian.m
...
Fix description of algorithm and document the function better. Also removes manual increasing of loop variable and moves computation of temporary variable not depending on loop variables outside of the loop as it needs to be computed only once.
2014-11-14 16:33:34 +01:00
Jukka Heikkonen
83931dca66
Added legends for the figures.
2014-11-14 14:53:37 +01:00
Jukka Heikkonen
90487c03e0
The naming of the resulting SA restrictions files (Smirnov statistics plots) changed so that the file names shows the endogenous variable and shock name of interest. E.g. If the model file name is ls2003a.mod the file name of the plot may be
...
ls2003a_prior_irf_calib_y_VS_e_ys_SA_1 showing the endogenous variable name y and shock e_ys.
2014-11-14 14:51:54 +01:00
Houtan Bastani
f999f960f8
bug fix: add missing comment
2014-11-14 12:27:20 +01:00
Houtan Bastani
b14c10c1a2
Merge pull request #775 from JohannesPfeifer/sim1_imaginary
...
Take care of imaginary numbers in perfect foresight simulations
2014-11-14 12:24:27 +01:00
Stéphane Adjemian (Telemachus)
2300464cae
Merge branch 'simplex'
2014-11-14 11:07:58 +01:00
Michel Juillard
9fa92b0d02
fixing error message
2014-11-14 07:15:24 +01:00
Stéphane Adjemian (Telemachus)
a8e7c30e6a
Added @dseries/exist method.
2014-11-13 16:53:06 +01:00
Johannes Pfeifer
cc96d46911
Transfer hard-coded verbosity of simplex algorithm into an option
2014-11-10 19:54:44 +01:00
Johannes Pfeifer
7c59edead2
Make simplex_optimization_routine.m usable outside of estimation
...
Removes hard-coded parameter names extracted from bayestopt_ by handing them over explicitly
2014-11-10 19:53:55 +01:00
Johannes Pfeifer
c3467ddca2
Simplex: return only scalar best value instead of full vector
2014-11-10 19:41:50 +01:00
Houtan Bastani
7b90fd9c32
Merge pull request #777 from JohannesPfeifer/simul_no_varexo
...
Add unit test for deterministic simulations without exogenous variables
2014-11-10 18:32:49 +01:00
Johannes Pfeifer
201bbccae9
Add unit test for deterministic simulations without exogenous variables
2014-11-10 18:25:01 +01:00
Stéphane Adjemian (Telemachus)
2e776d544a
Merge branch 'master' of github.com:DynareTeam/dynare
2014-11-10 15:17:54 +01:00
Stéphane Adjemian (Telemachus)
1285e0090d
Do not use git protocol to connect to github.
...
For submodules use https protocol (dynbot is not recognized by github).
2014-11-10 15:11:53 +01:00
Stéphane Adjemian (Sedna)
bbf6ea50b0
Added workaround for bug in some Matlab versions (wrong type).
2014-11-10 11:21:05 +01:00
Stéphane Adjemian (Telemachus)
0691b08303
Adapated internals routine (name of dynTest routine has changed).
2014-11-09 22:20:43 +01:00
Stéphane Adjemian (Telemachus)
6557849b30
Updated m-unit-tests module.
2014-11-09 22:20:18 +01:00
Stéphane Adjemian (Telemachus)
a12e9d684a
Force type consistency when comparing actual results and expected
...
results in unit tests (use logical type).
2014-11-09 22:17:46 +01:00