Commit Graph

9534 Commits (ee78ad20498f20179a28f9867ab607201b74ed16)

Author SHA1 Message Date
Stéphane Adjemian (Hermes) ee78ad2049 Improved execution speed.
Removed some calls to bsxfun which is not optimal on sparse matrices. I
only removed the calls where I identified bottlenecks (with a Smets and
Wouters model), more tests are needed to decide if we have to remove other
occurences of bsxfun on sparse matrices.
2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Hermes) be0f6230ad Cosmetic changes.
- Use a switch-case block.
 - Added a warning when the user tries to solve a linear(ized) model
with solve_algo different from 0 (for a linear model).
 - Added an error message if the user try to solve a linearized model
with stack_solve_algo=6 (not implemented).
2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Hermes) 1815eb8e56 Code factorization. 2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Hermes) 8a163842d0 Removed oo_ from sim1_linear routine. 2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Hermes) 7eed391eb7 Do not display max. abs. error. 2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Hermes) 4e8fcf6d47 Use lmmcp option in integration tests. 2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Hermes) 97d094b6dd Added mention to equation tags in Ref. manual. 2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Hermes) 53747cb082 Fixed lmmcp option.
This option may be used in extended_path and perfect_foresight_solver commands.
2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Hermes) 7355214413 Added integration tests for lmmcp in testsuite.
Also merged rbcii.mod and rbciia.mod.
2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Hermes) 8b084595e2 Added solve_algo as an option for perfect_foresight_solver. 2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Charybdis) 8763c29f34 Renamed and changed integration test for sim1_purely_forward routine.
Check the results against analytical solution. Also added a new
integration test (ar1.mod).
2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Charybdis) 02c71eb270 Renamed and changed integration test for sim1_purely_backward routine.
Compare the solution returned by sim1_purely_backward with the true solution.
2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Charybdis) e0be60710c Removed oo_ from sim1_purely_{backward,forward} routines. 2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Charybdis) 0abb9dc6f9 Added integration test.
Comparison of the solutions for the endogenous variables in perfect
foresight models with stack_solve_algo equal to 0 and 6 (LBJ, legacy
algorithm).
2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Charybdis) 290ce5db68 Removed oo_ from sim1_lbj routine. 2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Charybdis) 674c9ab286 Removed oo_ from sim1 routine. 2016-05-24 17:26:06 +02:00
Stéphane Adjemian (Hermes) 4658d81740 Add pointer to user written steady state routine. 2016-05-23 13:48:19 +02:00
Stéphane Adjemian (Hermes) f0229f3b1c Removed workspace() in examples. 2016-05-23 13:48:19 +02:00
Stéphane Adjemian (Hermes) 9c1739520b Bug fix in assert (Julia output)
Same as 4b830d29ed2d230908a1a5ec1c7badd29c7dc86a but for static model.
2016-05-23 13:48:19 +02:00
Stéphane Adjemian (Hermes) 0398832e4f Assign parameter values in main julia module.
See PR #1102 by spencerlyon2.

Added support for filling in the model.params Vector.
2016-05-23 13:48:19 +02:00
Stéphane Adjemian (Hermes) cf9883ef61 Bug fix in assert (Julia output).
See PR #1102 by spencerlyon2.

size always returns a tuple, and we were comparing to an int. Dispatch restricts these arguments to all be vectors so length works here.
2016-05-23 13:48:19 +02:00
Stéphane Adjemian (Hermes) e65c4f970b Fixed indexing of params in steady state file. 2016-05-23 13:48:19 +02:00
Stéphane Adjemian (Hermes) e44d90adc1 Fixed type issues.
params, exo_steady_state and steady_state are vectors not matrices!
2016-05-23 13:48:19 +02:00
Stéphane Adjemian (Hermes) 2e8f528bef Cosmetic changes.
+ Renamed model as model_
+ Renamed options as options_
+ Renamed output as oo_
+ Also export oo_ and options_
+ Fixed typos in comments.
2016-05-23 13:48:19 +02:00
Stéphane Adjemian (Hermes) a858ef6986 Updated .gitignore file. 2016-05-23 13:48:19 +02:00
Stéphane Adjemian (Hermes) c3b5aee714 Specialize @dynare macro if one model is processed.
In this case importall is used instead of import.
2016-05-23 13:48:19 +02:00
Stéphane Adjemian (Hermes) bcd39ef2f7 Made room for user written steady state file. 2016-05-23 13:48:19 +02:00
Stéphane Adjemian (Hermes) 96d538469c Added boolean variable for tracking if a steadystate block is present in the mod file. 2016-05-23 13:48:19 +02:00
Stéphane Adjemian (Hermes) 73787c0c16 Removed try-catch block. 2016-05-23 13:48:19 +02:00
Stéphane Adjemian (Hermes) 928b0a50a5 Fixed crash when steadystate block does not exist. 2016-05-23 13:48:19 +02:00
Stéphane Adjemian (Hermes) 4bf436ec33 Fixed typo in comments. 2016-05-23 13:48:19 +02:00
Stéphane Adjemian (Hermes) 5998558c32 Fixed deprecated require issue + Changed interface
+ Fixed deprecated require().

+ Changed @dynare macro and added new macro @compile.
  The new macro @compile just runs the preprocessor, while the macro
  @dynare calls compile and put the generated objects in the
  workspace. Note that this macro can take an arbitrary number of mod
  files as argument.

      *Example*

      @dynare "rbc.mod" "nk.mod"

      Will run the preprocessor on both models, and the generated
      data (model descriptions, options, ...) are available in the
      workspace. For instance, to obtain the number of equations:

	julia> rbc.model.eq_nbr
   	6

   	julia> nk.model.eq_nbr
   	10

   	@compile "rbc.mod" "nk.mod"

   	Will run the preprocessor on both models, but will not load the
   	generated data in the workspace. This can be done as follows:

   	julia> using rbc

+ Changed examples (rbc1 has a steady state block but not rbc2), using
  new interface.
2016-05-23 13:48:19 +02:00
Houtan Bastani 25f6d81cb7 Merge pull request #1200 from rattoma/gsa
identification and slice
2016-05-23 11:59:26 +02:00
Marco Ratto 3e8dbcfc7e Set mode_files in such a way that it can be handled with key-value pair.
We need a single mode file that contains variable xparams = [x1 x2 x3] all multiple modes, one per column.
Fixes #1199
2016-05-22 23:15:18 +02:00
Marco Ratto d8fea3ff13 Correlations and measurement errors in identification:
1) Filter out measurement errors with error message that suggests to explicitly write measurement errors in model definition
2) allow identification checks with correlations, by switching to numerical derivatives
2016-05-22 23:10:01 +02:00
Johannes Pfeifer 389f43019d Put fields of OSR in M_ into workspace to restore compatibility with 4.4.3 2016-05-22 11:41:54 +02:00
Johannes Pfeifer d10f8b42d7 Fix bug in convert_dyn_45_to_44.m if bayestopt_ is empty 2016-05-22 11:41:54 +02:00
Johannes Pfeifer fb57ef453d Put all OSR unit tests into subfolder and add one for parameter bounds 2016-05-22 11:41:54 +02:00
Johannes Pfeifer 03563e012b Allow for pass OSR parameter bounds to optimizer 2016-05-22 11:41:54 +02:00
Johannes Pfeifer 9774d0544d Initialize fields of M_.osr to empty in global_initialization.m 2016-05-22 11:41:54 +02:00
Johannes Pfeifer 676182f550 Cosmetic fix to dynare_minimize_objective.m 2016-05-22 11:41:54 +02:00
Johannes Pfeifer d9e22a149c Fix preprocessor renaming error introduced in b486ae4964 2016-05-22 11:41:54 +02:00
MichelJuillard 3df1a1275f Merge pull request #1185 from JohannesPfeifer/bytecode_fixes
Bytecode fixes
2016-05-22 10:32:56 +02:00
MichelJuillard 27a90957f9 Merge pull request #1198 from JohannesPfeifer/ident
Add missing else-clause in simulated_moment_uncertainty.m
2016-05-22 10:31:39 +02:00
Johannes Pfeifer 13d42e05a5 Add missing else-clause in simulated_moment_uncertainty.m
Fixes crash in unit tests
2016-05-22 10:03:56 +02:00
Stéphane Adjemian 0521aa4e09 Merge pull request #1188 from JohannesPfeifer/slice_marco
Add slice and factorization of posterior samplers
2016-05-22 01:45:14 +02:00
Stéphane Adjemian c7db18b35c Merge pull request #1192 from JohannesPfeifer/small_fixes
Small fixes
2016-05-21 23:47:44 +02:00
Houtan Bastani 5fcd502f3c preprocessor: allow comments in parallel config file. closes #1194 2016-05-20 17:22:26 +02:00
Houtan Bastani 2ce34e3207 preprocessor: replace obj_var_ with M_.osr.variable_indices and optim_weights with M_.osr.variable_weights. #948 2016-05-20 15:53:49 +02:00
Houtan Bastani 5fdd24ab8b preprocessor: add warning if more than one osr_params statement is encountered in the .mod file. #948 2016-05-20 15:46:33 +02:00