5024af0099
extended path: fixed calibrated shocks
2016-05-31 12:10:12 +02:00
c14125d585
PATH: fixing bugs
2016-05-31 12:10:12 +02:00
76f008e4e4
PATH: remove calls to lcppath that doesn't work with perfect foresight models
2016-05-31 12:10:12 +02:00
d106b20b69
perfect foresight: set boundaries with PATH
2016-05-31 12:10:12 +02:00
ddb9c0657c
fixing linear complementarity problem
2016-05-31 12:10:12 +02:00
450d7b099a
fixing error status for PATH hook
2016-05-31 12:10:12 +02:00
11161063d4
removing debugging output
2016-05-31 12:10:12 +02:00
c49332df43
add support for PATH for steady state
2016-05-31 12:10:12 +02:00
c159dff19c
adding hook for PATH solver
2016-05-31 12:10:12 +02:00
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
66ecdbf368
Merge pull request #1206 from JohannesPfeifer/doc_fixes
...
Various documentation and cosmetic fixes
2016-05-30 19:58:15 +02:00
14c4a5c131
Fix copy and paste mistake in manual
2016-05-30 17:20:26 +02:00
cd128cb563
Document configuration file setting for Windows
...
Closes #337
2016-05-30 13:58:38 +02:00
96098a17d2
Cosmetic fix to set_all_parameters.m
2016-05-30 13:31:22 +02:00
b58c8274dd
Manual: clarify that name tags are not required for all equations for display of residuals with tags
2016-05-28 19:34:07 +02:00
118b3ccb74
catstruct.m: remove call to nargchk, which will be removed in future Matlab versions
2016-05-28 19:32:11 +02:00
eca30eff75
Modify lmmcp unit test to have second (name) tag
2016-05-28 19:31:53 +02:00
f5fd699149
Update documentation on equation tags and LMMCP
...
Also documents solve_algo=10
2016-05-28 18:02:57 +02:00
06ed7ad83f
Update manual on slice mode_files option to reflect 3e8dbcfc7e
2016-05-28 15:45:39 +02:00
96ca17623b
Fix URL fir equation tags
2016-05-28 15:25:03 +02:00
dfbad0404d
Fixed crash in sim1_linear when periods=1.
...
Closes #1176 .
2016-05-26 22:19:36 +02:00
adb1471a3a
Added example for dynare++ in windows package.
...
Closes #1178 .
2016-05-26 15:05:59 +02:00
03da338b05
Added more informative error messages.
2016-05-25 12:54:47 +02:00
6007d51ef4
Code factorization.
2016-05-25 12:42:56 +02:00
fb4dfc6da6
Fixed typo.
2016-05-24 22:32:02 +02:00
0ea9bc4e90
Fixed typo introduced in 0abb9dc6f9.
2016-05-24 21:18:13 +02:00
cd55ed749c
extended_path: fixing inititalization for PATH
...
(cherry picked from commit 7d29e917f77e7e20211e5be01544d86c64af8c65)
2016-05-24 18:08:33 +02:00
28c0c92ae8
fixing typos
...
(cherry picked from commit 86995a3bd478a3dc02919424aeb77e2a550a84c5)
2016-05-24 17:57:26 +02:00
3fa6a96bfb
Added homotopy.
...
Fixes the issue of paths containing spurious solutions for (stochastic)
perfect foresight models.
2016-05-24 17:28:24 +02:00
cfc6957625
Cosmetic changes.
2016-05-24 17:28:24 +02:00
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
80f1baad7e
Changed solver in simul_backward_nonlinear_model.
2016-05-24 17:28:24 +02:00
6c7f3f1932
Added input to pass sequence of shocks.
2016-05-24 17:28:24 +02:00
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
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
fbf2fae036
Changed option value for unscented integration nodes.
2016-05-24 17:28:24 +02:00
edce6b4779
Added integration test for Monte-Carlo EP.
2016-05-24 17:28:24 +02:00
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
3bbac629ed
Cosmetic change.
...
Removed maximum_lead and maximum_lag in extended_path routines.
2016-05-24 17:28:24 +02:00
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
6dcfd1c072
Added integration test.
...
Test that the extended path and stoch_simul (with order equal to 1)
algorithms return the same paths for the endogenous variables if the
RE model is linear.
2016-05-24 17:28:24 +02:00
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
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
e507a6d0a9
Fixed copyright header.
2016-05-24 17:28:24 +02:00
8958b7ef98
Cosmetic changes.
2016-05-24 17:28:24 +02:00
3a1f74d7a7
Fixed typo.
2016-05-24 17:28:24 +02:00
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
b60bd7b36b
Removed globals from extended_path routine.
2016-05-24 17:28:24 +02:00
635d5b704b
Added description of options lmmcp and stack_solve_algo==7 in reference manual.
2016-05-24 17:28:24 +02:00
4bd8ef7a9f
Added new integration tests.
...
Simulation of the Smets and Wouters perfect foresight model, with a
productivity shock such that the nominal interest rate hits the
ZLB. Comparison of the solutions returned by a Newton
algorithm (stack_solve_algo==0) and the LMMCP algorithm. AT the time of
this commit, the results are different... Probably an issue with the
LMMCP algorithm.
2016-05-24 17:28:24 +02:00
Michel Juillard
Stéphane Adjemian
Johannes Pfeifer