Ferhat Mihoubi
d586280b1b
- The binary operator oDerivPower is implemented in bytecode
2010-12-17 18:54:03 +01:00
Ferhat Mihoubi
7a6776197d
- The function getPowerDeriv must end with an "end" instruction
2010-12-17 18:54:02 +01:00
Houtan Bastani
1dd5493c1a
parallel: removed warning and clarified error message
2010-12-17 12:16:27 +01:00
Houtan Bastani
c7e5b44801
parallel: SingleCompThread bug fix
2010-12-17 12:03:37 +01:00
Houtan Bastani
7a26fe2ebd
extend domain of derivative of x^p to x=0 under certain conditions (see ticket 78)
2010-12-13 17:03:57 +01:00
Sébastien Villemot
8af9e0dd95
Preprocessor: no longer use boost::graph::GraphvizDigraph class, removed in Boost 1.44 ( closes #155 )
2010-12-13 14:23:04 +01:00
Ferhat Mihoubi
7dc09890fb
- Bug correction during the block decomposition: the models' graph was ill oriented before the search of strong components
2010-12-10 15:11:12 +01:00
Ferhat Mihoubi
751c0da2a1
- external functions are compatible with block decomposition and/or bytecode
2010-12-10 12:06:50 +01:00
Michel Juillard
d0b7e9e59f
added choice of parameter set to shock_decomposition. This corrects a bug when shock_decomposition is requested after Metropolis iterations.
2010-12-08 13:02:59 +01:00
Houtan Bastani
d206217fef
fixed bug that would force node sharing when two or more expectation operators had the same expression but different information sets
2010-12-08 12:29:47 +01:00
Sébastien Villemot
405ca1365c
Preprocessor: fix bug with varexo_det in conjunction with USE_DLL
2010-12-08 11:33:20 +01:00
Sébastien Villemot
818f7df793
Preprocessor: delete remaining dynamic.m file when using USE_DLL
2010-12-08 11:32:13 +01:00
Sébastien Villemot
98168dabd8
MOD files ending with @#endif/@#endfor but no newline now work ( closes #73 )
2010-12-06 17:59:30 +01:00
Sébastien Villemot
2984f61fe0
Bugfix at order 3 (hopefully was not affecting final results, thanks to Pablo Winant for spotting this):
...
- in the 3rd derivatives matrix, among symmetric elements, the first one had
the right values but the following ones were set to zero
- moreover, the k-order DLL was trying to add all the symmetric elements in the
folded tensor, instead of only keeping one value among all the symmetric ones
- hopefully, Ondra's tensor library was (silently) refusing to add symmetric
elements after the first (and right) value had been added
- so the final result was correct
2010-12-06 17:00:46 +01:00
Ferhat Mihoubi
cf3967d19d
- Correction of a bug in the Jacobian Matrix computation for a non block-decomposed model using bytecode
2010-11-25 16:11:22 +01:00
Ferhat Mihoubi
3f6ddd1c3f
- Adds the derivatives with respect to deterministic exogenous variables
...
and corrects the computation of the number of column in the jacobian matrices in a block decomposed model without bytecode
- Identical corrections for models using bytecode without block decomposition
2010-11-25 16:11:12 +01:00
Ferhat Mihoubi
5702815f97
Correction of bugs in check command with block decomposition and
...
Add the missing dr_block m-file
2010-11-25 16:10:46 +01:00
Sébastien Villemot
33d340dd04
Preprocessor: also perform numeric initializations in stochastic context
2010-11-25 16:07:55 +01:00
Sébastien Villemot
eaad0b1263
Preprocessor: various cleanups related to the signed vs non-negative numerical constants issues
2010-11-25 13:45:35 +01:00
Sébastien Villemot
973c795db8
Change the syntax for values of deterministic shocks:
...
Arbirtrary expressions after the "values" keywords must now be enclosed
within parentheses; consider the following example:
<code>
periods 1:2;
values -1 -2;
</code>
In the previous syntax, this was interpreted by the preprocessor as a shock of
value -1-2 = -3 for periods 1 and 2, which is clearly not the intent of the
user; with the new syntax, this will be rejected (too many values compared to
the number of ranges).
Also note that now commas are no longer required between arbitrary expressions,
since the parentheses are sufficient for separating them.
2010-11-24 18:26:43 +01:00
Sébastien Villemot
0269c5766c
Preprocessor: max lead/lags in the model are no longer affected by lead/lags on parameters ( closes #153 )
2010-11-18 11:21:20 +01:00
Sébastien Villemot
acce150795
New "console" option ( closes #130 )
2010-10-28 11:29:39 +02:00
Houtan Bastani
ad73ab0a83
Dynare parallel: preprocessor implementation
2010-10-28 10:54:43 +02:00
Ferhat Mihoubi
b781c44e5a
Check command is now compatible with block and bytecode options
2010-10-27 15:34:48 +02:00
Ferhat Mihoubi
ab311cc0fd
- Store the incidence matrix containing reordered equations and variable in block_structure
2010-10-27 15:30:01 +02:00
Ferhat Mihoubi
c89682670d
- The Jacobian matrix in the block decomposed model without bytecode option is now sparse
2010-10-22 16:20:28 +02:00
Ferhat Mihoubi
f7f83d6973
- Correction of several bugs with Octave
...
- Add wrapper needed to compute the steady-state using bytecode and solve_algo = 0, ..., 4
2010-10-22 10:09:07 +02:00
Sébastien Villemot
e9474ae9cc
Histval: fixed bug for exogenous introduced in previous commit
2010-10-20 14:56:35 +02:00
Sébastien Villemot
615f2ae322
Histval block:
...
- fail when user tries to initialize a variable at a period which is not in the model, and make the error message explicit
- correctly handle exogenous variables in stochastic mode, which are also substituted with an aux var
- added test files
2010-10-20 14:47:03 +02:00
Sébastien Villemot
855fdf8c16
Preprocessor: fix possible bug in histval (need to limit search of aux vars to aux vars for lagged vars, otherwise we could have a wrong match with some other kind)
2010-10-20 11:39:09 +02:00
Houtan Bastani
a7fc3361ac
trend_var, growth_factor, deflator: stationarize a nonstationary model and test
2010-10-18 16:07:42 +02:00
Houtan Bastani
0782dd57e3
add preprocessor simplification rule
2010-10-18 15:51:07 +02:00
Ferhat Mihoubi
810416ebe6
- Extends the algorithms compatible with bytecode to compute the steady-state (ticket #11 ). The new values for solve_algo are:
...
* 0: fsolve
* 1: solve1
* 2, 4: solve1 + block decomposition
* 3: csolve
* 5: LU decomposition with UMFPack (method handling sparse matrix in Matlab)
* 6: GMRES
* 7: BiCGStab
* 8: bytecode own solver (use Gaussian elimination + sparse matrix)
- Bytecode can now evaluate a specific block instead of the overall blocks (new bytecode's option 'Block = block_number')
2010-10-11 19:21:32 +02:00
Sébastien Villemot
d973c2ead7
Preprocessor: give more explicit error message when lead or lag is used inside 'planner_objective'
2010-10-11 14:49:23 +02:00
Sébastien Villemot
3218d3c72f
Preprocessor: don't create auxiliary variables for expressions appearing only
...
in unused model local variables
2010-10-11 12:58:38 +02:00
Houtan Bastani
5327f161d4
external_function: add support for use_dll
2010-10-07 18:05:07 +02:00
Houtan Bastani
85564b287c
external_function: modified to pass functions as string names as opposed to function handles to reduce number of calls to mexCallMatlab when use_dll option is invoked
2010-10-07 16:49:36 +02:00
Houtan Bastani
a57aa4359a
replace deprecated Matlab function strvcat with char
2010-09-27 16:11:14 +02:00
Ferhat Mihoubi
83648a6d53
- Correct the simul call without input argument
2010-09-24 11:07:43 +02:00
Sébastien Villemot
376e25ca67
Preprocessor: no longer crashes when some dynamic variables appear only in
...
unused model local variables (closes #101 )
2010-09-23 12:13:30 +02:00
Sébastien Villemot
0e73523f3b
Removed all MSVC workarounds, since we can now definitely drop this compiler on
...
Windows/64-bit platform
2010-09-21 15:06:14 +02:00
Sébastien Villemot
63d0fb3f04
Preprocessor: implement STEADY_STATE operator with "use_dll" and "block"
...
options (closes #98 )
2010-09-20 17:04:38 +02:00
Ferhat Mihoubi
5490f1bb09
RESID command is compatible now with block option and without bytecode option (closes ticket #45 )
2010-09-17 16:54:43 +02:00
Ferhat Mihoubi
f026137fa9
- Adding the relaxation method for block and/or bytecode options
...
- Changing the the stack_solve_algo option :
Previous version New version
1 : Sparse LU 0 : Sparse LU
2 : GMRES 1 : Relaxation
3 : BiCGStab 2 : GMRES
4 : Sparse LU & Optimal Path 3 : BiCGStab
5 : bytecode own solver 4 : Sparse LU & Optimal Path
5 : bytecode own solver
2010-09-17 12:18:25 +02:00
Ferhat Mihoubi
3e1a2ad450
- Substitutes lead for exogenous and endogenous variables in the deterministic case: direct substitution of the variables even in non-linear expressions
2010-09-17 12:17:09 +02:00
Sébastien Villemot
ffd2cb24aa
Preprocessor: make option "no_static" work even when "bytecode" is not present
2010-09-17 12:17:05 +02:00
Ferhat Mihoubi
2da9d5adda
Bug correction in the deterministic simulation: to solve a purely backward or forward looking block requires contemporaneous derivatives only.
2010-09-17 12:17:00 +02:00
Ferhat Mihoubi
f68e4fd455
- compute the jacobian matrix for exogenous, det_exogenous and previous blocks endogenous
...
- preprocessor provides informations about the dynamic structure of each block
- extends the algorithms available with bytecode: stack_solve_algo = 1, 2, 3 and 4 is compatible the bytecode. Speed and memory requirement are improved with stack_solve_algo = 1 or 4 for large scale models.
- bytecode can be used to evaluate a model
2010-09-17 12:16:47 +02:00
Sébastien Villemot
2c10d03368
Preprocessor: rename "NodeID" in "expr_t"
2010-09-17 12:13:26 +02:00
Sébastien Villemot
c8fdd5d9d1
Preprocessor: uniformize typedef names (with a "_t" suffix)
2010-09-16 19:00:48 +02:00