064a3dcd77
Fix orig_lead_lag field for unary op auxiliary variables
...
Ref. dynare#1843
2022-01-19 15:40:24 +01:00
d149d3a76c
PAC: substitute unary ops in “growth” expressions (in “pac_model” and “pac_target_info”)
...
Ref. dynare#1837
2022-01-18 12:40:15 +01:00
f05b552f31
PAC: correctly handle log() and diff() operators in the equation defining the target
...
By default, the preprocessor substitutes logs and diffs by auxiliary variables
only in equations of auxiliary models (VAR, PAC). Hence they are not
substituted in the equation defining the target, since it is outside the
auxiliary model.
The fix consists in doing the substitution in that equation as well, when we
parse the linear combination within that equation.
Ref. dynare#1837
2022-01-17 12:57:37 +01:00
8c528f4f9a
Add missing include
...
(“endl” is defined in “ostream”)
2022-01-06 14:55:57 +01:00
0db05a886a
Move “using namespace std” statement after inclusion of standard headers
...
Otherwise clang emits this warning:
using directive refers to implicitly-defined namespace 'std'
2022-01-06 14:54:57 +01:00
75a26b6d11
Remove unused member in the RamseyPolicyStatement class
2022-01-06 14:47:34 +01:00
5572e478f1
Indentation fix
2022-01-06 14:46:38 +01:00
ebf7b6c51d
Clang compatibility fix
2022-01-06 14:35:39 +01:00
a8fce06dc4
PAC: new options “auxname” and “kind” to “pac_model”
2021-12-16 15:40:47 +01:00
ea6fb40db7
PAC: new “pac_target_info” block and “pac_target_nonstationary” operator
...
Ref. Madysson/estimation-codes#5
2021-12-16 15:40:47 +01:00
6a31ba4b62
PAC: minor simplification in the matching of the growth expression
...
Simple variable nodes are already correctly matched by
ExprNode::matchLinearCombinationOfVariables().
Also improve the related documentation.
2021-12-16 15:40:47 +01:00
7dde09169e
estimated_params: check that no parameter is declared twice across concatenated blocks
...
The existing check (in AbstractEstimatedParamsStatement::commonCheckPass()) was
only working inside a given estimated_params block, but not across blocks.
2021-12-16 15:22:48 +01:00
ea44aa19d6
New option “overwrite” to “estimated_params” block
...
This option provides the same behaviour as before the change in
60ef6bbdbd
2021-12-16 14:22:31 +01:00
5ffbc5bad3
New estimated_params_remove statement
2021-12-15 16:02:57 +01:00
06a436fdf0
Minors simplification in EstimatedParams* classes
...
Also use a better name for the type-specific ID variable.
2021-12-15 16:02:53 +01:00
342c4faf8d
New var_remove statement
2021-12-15 12:20:29 +01:00
60ef6bbdbd
Allow multiple estimated_params statements
2021-12-15 12:20:29 +01:00
2d07186b8b
New model_options statement
...
The purpose of this statement is to specify the options that apply to the whole
model, when there are several “model” blocks, so as to restore the symmetry
between those blocks (since otherwise one “model” block would typically bear
the options, while the other ones would typically have no option).
It still remains possible to specify the options on one “model” block, or even
spread accross several ones.
Closes : #19
2021-12-15 12:20:14 +01:00
1f004584e9
New model_remove statement and model_replace block
...
Factorize the implementation with the existing code for the include_eqs and
exclude_eqs options (which has been refactored by the way).
Ref. #19
2021-12-15 12:11:25 +01:00
a73b0d911a
Fix logic of include_eqs in the presence of [static]/[dynamic] equations
...
In a model with [static]/[dynamic] equations, if the user was using include_eqs
with a list of equations that does *not* contain equations marked
[static]/[dynamic], then the call to ModelTree::includeExcludeEquations(…,
static_equations=true) would have an empty list of equation tags (as stored
in tag_eqns).
The right behaviour in this case is to exclude all static equations. However,
the code would exclude none, and this would disrupt the equilibrium between
[static] and [dynamic] equations (since all [dynamic] equations were excluded
by the other call to the same method).
The fix consists in removing the shortcut that returns from the method if
tag_eqns is empty.
2021-12-10 12:50:10 +01:00
a492417a5b
Fix typo
2021-12-10 12:44:42 +01:00
d047878333
Julia specific interface for normcdf and normpdf
2021-12-10 10:17:06 +01:00
c0ea8d7203
Add hyperbolic primitives (cosh, sinh, tanh, acosh, asinh, atanh)
...
Everything was already in place (since ages!), except that the parser interface
was missing.
Also fix the derivation formula for atanh, which was incorrect.
2021-12-07 18:03:33 +01:00
d15b2110a0
Add erfc() primitive
...
Closes : #85
2021-12-07 15:19:40 +01:00
a040a7dbde
Merge branch 'write_regimes' of git.dynare.org:JohannesPfeifer/preprocessor
...
Ref. !55
2021-12-07 12:24:38 +01:00
81abea042b
Occbin: use element-by-element multiply, divide and power in occbin_difference.m
...
This file operates on vectors.
Ref. dynare#1826
2021-12-07 12:22:59 +01:00
b356b8d5fa
Substitute out adl() operators in equations marked [static]
2021-12-06 17:17:38 +01:00
dc99d81c45
Explain why expectation() operators need not be substituted in equations marked [static]
...
By the way, also perform a small code simplification.
2021-12-06 17:17:38 +01:00
214f0db4db
Add comment about predetermined variables and equations marked [static]
...
By the way, also perform a small code simplification.
2021-12-06 17:17:38 +01:00
5fe94ed606
Simplify constant equations also in equations marked [static]
...
The simplifyEquations() methods had to be moved to DynamicModel, in order to
access the static_only_equations member.
2021-12-06 17:17:37 +01:00
fe974c4c31
When looking for unused endos/exos, also search in equations marked [static]
2021-12-06 17:17:37 +01:00
8d4cab6fe3
Substitute model-local variables in equations marked [static]
2021-12-06 17:17:27 +01:00
164e2debdc
occbin_write_regimes: add option to select type
2021-12-06 16:10:58 +01:00
7b8fc8edb7
Symbolic detrending engine: correctly handle equations marked [static]
...
Incidentally, this also fixes a detrending bug in the Occbin engine (since the
latter internally generates a [static] equation).
Ref. dynare#1827
By the way, perform a small code simplification.
2021-12-06 12:59:21 +01:00
607c5aab19
Merge branch 'planner_objective_options' into 'master'
...
evaluate_planner_objective: add options
See merge request Dynare/preprocessor!54
2021-12-03 16:07:14 +00:00
befd432be7
evaluate_planner_objective: add options
...
Related to https://git.dynare.org/Dynare/dynare/-/merge_requests/1967
2021-12-03 14:06:57 +01:00
3c6d5126b3
PAC: fix comment
2021-12-03 11:07:59 +01:00
a210a8fd59
Add several types of auxiliary variables to M_.mapping
...
In practice, only those auxiliary variables which do not have an orig_symb_id
will be listed (in addition to unary ops, due to an implementation bug).
2021-11-24 11:27:17 +01:00
2282d4773c
PAC: new auxiliary variable type for pac_expectation operator
...
Also add the new equation to auxiliary equations in the backward case (but not
in the MCE case, since that equation is recursive).
2021-11-23 14:31:03 +01:00
ae242770f9
PAC: in the backward case, unify h0 and h1 parameters vector into a single vector
...
There is no reason to keep this distinction.
Additionally, since the data structure is now symmetric with the MCE case,
unify this “h” parameter vector with the “α” parameter vector.
2021-11-23 14:31:03 +01:00
0f89569f7e
PAC: in the backward case, create an aux. var. for the pac_expectation expression
...
By the way, refactor the MCE case by merging two routines related to the Z₁ aux. var.
This restores the symmetry with the backward case, now that the latter also has
an aux. var. for the pac_expectation operator. Also store the aux. var. IDs in a
structure common to the backward and MCE cases.
2021-11-23 14:31:03 +01:00
0e62206203
PAC: drop support for multiple PAC equations associated to a given PAC model
2021-11-23 14:31:02 +01:00
1cc512962c
No longer consider an equation of the form “x(+1) = 0” as defining x to be a constant
...
Fix this by restricting the detection of constant equations to those where the
variable appears without a lead or lag.
Closes : #83
2021-11-19 18:01:11 +01:00
b51f122e59
With nostrict, really ignore unused endogenous in initval/endval/histval
...
Closes : #82
2021-11-19 17:49:57 +01:00
02ee077663
Fix handling of undeclared variables in the planner_objective statement
...
Closes : #81
2021-11-19 17:32:45 +01:00
1e4eadb266
Macroprocessor: make the removal of “@#line” directives and extra newlines consistent across LF and CR+LF conventions
...
Simplify the algorithm using only regexps, and add comments.
2021-11-19 16:49:21 +01:00
1de3476546
Macroprocessor: fix line counter inside @{} constructs
...
There were two bugs:
– with the LF convention, newlines were counted twice
– with the CR+LF convention, they were counted four times (because the CR was
included in yyleng, alongside the LF)
The fix consists in implementing a location_increment() method similar to the
one used for the Dynare parser. This is the most robust solution.
By the way, mark the method DynareFlex::location_increment() method static.
2021-11-19 14:55:14 +01:00
487c372900
Macroprocessor: fix newlines in output when under Windows and input uses CR+LF convention
...
If the input .mod file uses CR+LF convention, and if the user is under Windows,
then the output of the macroprocessor (as given by the “savemacro” option) had
incorrect end of lines: those would be CR+CR+LF.
The reason is that some TextNode(s) internally created by the macroprocessor
would themselves contain CR+LF sequences, which would then be transformed into
CR+CR+LF in the output (because MinGW transforms LF into CR+LF in output
streams).
The fix consists in changing the nature of the EOL token: the parsed text is no
longer attached to it, so that the Bison file now systematically turns it into
a LF inside TextNode(s).
Closes : #80
2021-11-19 14:55:12 +01:00
1041f205ea
PAC: trap a few more cases when an auxiliary param conflicts with a declared one
2021-11-09 15:48:59 +01:00
6ad3d02ec6
Minor simplification in ExprNode::matchLinearCombinationOfVariables()
2021-10-29 15:39:40 +02:00
c4e9432eef
Remove implementation of various methods of PacExpectationNode
...
Those implementations are dummy and are normally never used. Rather error out
on those, factorizing them within SubModelNode.
2021-10-29 14:11:30 +02:00
3e242dd481
New class SubModelNode to factorize between VarExpectationNode and PacExpectationNode
2021-10-29 14:11:30 +02:00
37a6ed6261
PAC: rethink the relation between PacModelTable and DynamicModel classes
2021-10-29 14:11:30 +02:00
127361971f
PAC: refactor with new PacModelTable that replaces PacModelStatement(s)
...
Incidentally, the auxiliary endogenous variable representing Z₁ and created for
PAC MCE models no longer necessarily appear as the first auxiliary variable (so
this is effectively a revert of 64f55e4a5e
2021-10-29 14:11:30 +02:00
0123326123
Fix error message when using pac_expectation with occbin_constraints
...
By the way, also fix the argument name (the model is not a VAR).
2021-10-29 14:11:08 +02:00
bc827326eb
JSON: fix syntax error in the aux vars table introduced in a7fb198463
2021-10-26 12:31:01 +02:00
5bef185370
PAC: add growth correction term for definition of Z₁ in the MCE case
2021-10-25 17:28:01 +02:00
4191cec9aa
Merge branch 'master' into 'master'
...
Master
See merge request Dynare/preprocessor!53
2021-10-08 13:42:16 +00:00
8bb8e59410
Improve message for GCC installation under macOS
2021-10-08 12:59:17 +02:00
a7fb198463
add field aux_vars to modfile.json
2021-10-08 09:26:49 +02:00
561bee0790
add NNZderivatives to JSON + use AbstractVector for Julia functions
2021-10-06 17:45:11 +02:00
78e295115a
PAC: better error message in case of name conflict for the growth neutrality correction parameter
2021-10-06 16:41:09 +02:00
b496c4ba2c
ramsey_model: Fix error message logic
...
one of the statements being present is sufficient
2021-09-24 09:14:41 +02:00
eb006ecaeb
Remove unused steady_state_growth option to pac_model
2021-09-22 16:26:42 +02:00
8d14e18cdb
Rename mh_mode to correct mh_posterior_mode_estimation option
...
Closes https://git.dynare.org/Dynare/preprocessor/-/issues/60
2021-09-21 17:28:37 +02:00
b12746190d
ramsey_{model,policy}, discretionary_policy: check that instruments are declared endogenous
...
Closes : #72
2021-09-21 17:17:36 +02:00
3d215020ae
Fix use_dll on macOS + Octave
...
Enforce GCC with mkoctfile. Otherwise Clang will be used, and it does not
accept our custom optimization flags.
Ref. dynare#1797
2021-09-13 16:50:13 +02:00
989347f448
Fix bug introduced in previous commit that triggers a crash for DSGE-VAR models
2021-09-08 19:00:15 +02:00
cb19ece049
estimated_params*: add a check that prevents (apparently) recursive prior definitions
...
Closes : #77
2021-09-08 17:10:56 +02:00
62ba9a434a
estimated_params_{bounds,init} now check that a parameter (or correlation) is never declared twice
...
This check was previously only implemented in estimated_params.
2021-09-08 16:38:15 +02:00
784dd4122a
PAC: generate the “growth_neutrality_param_index” MATLAB field for MCE models
...
Previously it was only generated in the case of backward PAC models.
2021-08-31 17:54:20 +02:00
6863494db2
Ensure that the “discount” option of “pac_model” is a parameter
...
Ref. Madysson/estimation-codes#4
2021-08-31 12:52:00 +02:00
15658fda5c
Merge branch 'mom_namespace' of git.dynare.org:JohannesPfeifer/preprocessor
...
Ref. !50
2021-08-30 16:05:44 +02:00
297888ce72
Merge remote-tracking branch 'origin/updated_covariance'
...
Ref. !47
2021-08-30 16:05:13 +02:00
Stéphane Adjemian (Charybdis)
79d5a2d687
Fix warning (missing option in occbin_setup).
2021-08-18 15:18:21 +02:00
3f37d9692c
method_of_moments: move to namespace
2021-08-15 20:44:24 +02:00
6fb86522ea
Occbin: Add missing options
2021-08-10 20:28:32 +02:00
bb19d98712
Merge branch 'results_file' into 'master'
...
Move results file to output folder
See merge request Dynare/preprocessor!48
2021-07-23 11:58:46 +00:00
361b789e88
Move results file to output folder
2021-07-22 19:36:12 +02:00
38714a58c4
Add updated_covariance option
...
Related to https://git.dynare.org/Dynare/dynare/-/issues/1784
2021-07-22 17:12:02 +02:00
5cfe6303e2
Cosmetic changes
2021-07-22 16:49:59 +02:00
5da01faa74
Introduce option consider_all_endogenous_and_auxiliary
2021-07-22 11:49:36 +02:00
e1f7d8c735
VAR models: add a parameter representing the constant in the development of the pac_expectation operator
2021-07-21 17:53:55 +02:00
da9e0acc2d
VAR models: add a parameter representing the constant in the development of the var_expectation operator
2021-07-21 17:53:55 +02:00
1a5fb4d5e6
DynamicModel: make PAC data members private
2021-07-21 17:53:55 +02:00
Stéphane Adjemian (Ryûk)
8305b94d29
Fix constant definition in reduced form VAR.
2021-07-21 17:53:55 +02:00
Stéphane Adjemian (Ryûk)
c6a7d56b36
Fix constants in var model (it has to be a vector).
2021-07-21 17:53:55 +02:00
a6add6127b
VAR models: output vector of constants in varmatrices.m
2021-07-21 17:53:54 +02:00
1cf03a5f07
Cosmetic improvement
2021-07-21 17:53:54 +02:00
5498ce98ee
Use NumConstNode* type for DataTree constants (previously they were just expr_t)
2021-07-21 17:53:54 +02:00
15d7432105
Occbin: add more sanity checks on expressions in “occbin_constraints” block
...
— forbid leads and lags
— forbid expectation operators
This is implemented by moving the occbin_constraints expression to a separate
DataTree. As a side-effect, this removes the spurious non-linearity warning in
a stochastic context (because we were introducing inequalities and abs()
operators in the main DynamicModel tree).
2021-07-21 16:33:23 +02:00
ed31d1cb5e
Remove unused functions
2021-07-21 16:33:23 +02:00
6ae3d3131f
Occbin: handle external functions in “occbin_constraints”
2021-07-21 16:33:19 +02:00
0d6eb053d4
Ensure that “pac_expectation” operators are never inserted into a static model
...
The implementation of PacExpectationNode::toStatic() did not make sense. It
should normally never be called, hence we error out at this point.
Also do a cosmetic change in the PacExpectationNode::clone() methode.
2021-07-21 16:33:14 +02:00
b16ab9b247
Fix preprocessor crash when a [static] equation contains a “var_expectation” operator
...
Rather error out with a meaningful message.
2021-07-21 16:33:09 +02:00
ebd9954d33
Occbin: new syntax for constructing multi-regime equations
...
Ref. #68
2021-07-20 18:18:24 +02:00
0b8e3345fc
Ensure that a tag is not declared twice in the same equation
...
Previously the first value was silently ignored. It’s better to explicitly
forbid this. If this change creates problems (since it’s a change in an
undocumented behaviour), we can turn it into a warning.
2021-07-20 14:56:48 +02:00
81d4fd5d83
Occbin: generate +<basename>/occbin_difference.m
...
In particular, this implies that steady state values of endogenous in the
“occbin_constraints” block must now be specified using the STEADY_STATE()
operator (and not with a “_ss” suffix).
Moreover:
– make various simplifications to the fields generated under M_
– in the driver file, replace the call to occbin.initialize() by a few explicit operations
Ref. #68
2021-07-20 12:26:11 +02:00
90b1235a64
New utility to check if an expression contains an exogenous (possibly deterministic)
2021-07-20 12:10:58 +02:00
08ac455fcd
Implement “occbin_constraints” block
...
Consequently drop “occbin” option to “model”.
Incidentally, allow more values in equation tag names (previously some keywords
such as “alpha” were disallowed).
Ref. #68
2021-07-16 11:34:27 +02:00
Stéphane Adjemian (Ryûk)
6b9d94405c
Return the reduced form of the structural VAR model optionally.
...
Also
- changed the name of the generated matlab routine,
- changed initialization of matrix a0.
2021-07-16 08:33:26 +02:00
2e696a2e0d
Add shocks(surprise) block (for occbin)
...
Ref. #68
2021-07-13 21:55:31 +02:00
f0002c5ac6
Fix compiler warning introduced by 61b37942ab
2021-07-13 17:18:06 +02:00
c3b0ce8e69
shocks/mshocks/conditional_forecast_paths: various cleanups
2021-07-13 17:17:57 +02:00
5624260a07
Merge branch 'occbin_rebase' into 'master'
...
First part of Occbin interface
See merge request Dynare/preprocessor!42
2021-07-13 13:38:26 +00:00
61b37942ab
First part of Occbin interface
2021-07-13 13:38:25 +00:00
c9c36c037a
New preprocessor option “nocommutativity”
...
This option tells the preprocessor not to use the commutativity of addition and
multiplication when looking for common subexpressions.
As a consequence, when using this option, equations in various outputs (LaTeX,
JSON…) will appear as the user entered them.
There is however a potential performance cost to using this option, yet to be
determined.
Ref. dynare#1788
2021-07-13 14:35:10 +02:00
6aeef11bb2
New perfect_foresight_with_expectation_errors_{setup,solver} commands
2021-07-09 17:14:22 +02:00
365fb27f3d
New “structural” option to “var_model”
...
As the name implies, this option allows contemporaneous variables on the RHS.
The A0 matrix for contemporaneous variables is added as a second (optional)
output to the generated var_ar.m file. Note that for reduced-form VAR, this
matrix will be the identity.
Also, the user is now allowed to write the VAR models in a more flexible form:
the LHS must still be a single variable, but the RHS can be an arbitrary
expression (as long as it is linear, obviously). Internally, the preprocessor
now uses derivation to compute the coefficients of the AR and A0. This change
applies to both reduced-form and structural VAR models.
Ref. dynare#1785
2021-07-08 12:55:47 +02:00
f7a367d7b1
“var_model” statement: make “eqtags” option mandatory
2021-07-08 12:55:40 +02:00
67f4777ac3
Remove support for obsolete syntax: var_model(order = 2) X Y Z;
2021-07-07 18:26:09 +02:00
4cc22897b7
Cosmetic improvements
2021-07-06 19:09:25 +02:00
8df34c1ca5
VAR and trend component models: improve sanity checks on lead/lags on RHS
...
In particular, forbid lagged exogenous.
2021-07-06 18:52:38 +02:00
c9d6d9c6f8
Enforce non-positivity of the “time_shift” option to “var_expectation_model”
...
Closes : #70
2021-07-06 16:40:20 +02:00
221daeab6f
New time_shift option to var_expectation_model
...
Ref. dynare#1787
2021-07-01 17:56:07 +02:00
5a33699f03
VAR / trend component models: cosmetic changes
2021-07-01 17:52:05 +02:00
700dd31d57
Merge branch 'mom_varlist' into 'master'
...
method_of_moments: remove var_list_
Closes #69
See merge request Dynare/preprocessor!44
2021-06-28 09:32:07 +00:00
e6e25a5c87
macOS: fix detection of GCC installed by Dynare package
...
Since the preprocessor has moved, the relative path of the local Homebrew needs
to be adjusted.
2021-06-24 15:16:55 +02:00
f354458418
method_of_moments: remove var_list_
2021-06-24 15:10:24 +02:00
1c3e6f89af
macOS: bump to GCC 11 (for compilation and for use_dll)
2021-06-24 12:34:29 +02:00
824433aa3f
Refactor driver output for heteroskedastic_shocks
...
Ref. dynare#1791
2021-06-15 12:53:40 +02:00
d0ec5eb155
Allows to write M_.objective_tmp_nbr in the driver.m file
...
Writes the number of temporary terms of the planner objective function derivatives objective_tmp_nbr in the driver.m file.
Useful for the k-order approximations of welfare.
2021-06-11 12:37:31 +02:00
74ef3457b4
Use secure URL for link to GNU licenses
2021-06-09 17:21:00 +02:00
86b2b4c2da
When order option > 3, compute planner objective derivatives at that order
...
Ref. dynare!1866
2021-06-04 16:20:29 +02:00
04f242f971
xls_sheet option: allow quoted strings as value
...
Closes : #67
2021-05-31 18:18:25 +02:00
4b401e6253
Preparations for Occbin MR
...
Related to https://git.dynare.org/Dynare/dynare/-/merge_requests/1863
2021-05-31 15:12:46 +02:00
c5f9ae89e7
Implement new “occbin” option to “model” block
...
For the time being, the preprocessor will refuse that this option be used with
any command other than estimation.
By the way, remove occbin_likelihood and occbin_smoother options to estimation.
Ref. dynare#569
2021-05-31 10:00:59 +02:00
1e9ea1acf4
Merge branch 'smoother_redux' into 'master'
...
Add smoother_redux option
See merge request Dynare/preprocessor!38
2021-05-27 14:55:35 +00:00
2a8726fe6e
Add kalman_algo and filter_covariance to calib_smoother
2021-05-27 16:42:10 +02:00
0d635aed36
Add smoother_redux option
2021-05-27 16:41:36 +02:00
b12b254f2c
Add heteroskedastic_filter option and heteroskedastic_shocks block
...
Ref. dynare!1844, dynare!1851
2021-05-26 18:27:52 +02:00
e552bfc4b3
Refactor parsing of deterministic shocks
...
This will simplify the implementation of heteroskedastic_shocks.
2021-05-26 17:05:51 +02:00
c35f160e3f
Minor simplification
2021-05-26 16:17:18 +02:00
97e9609c28
Streamline use_dll compilation under MATLAB
...
— drop Windows 32-bit support
— drop -fno-omit-frame-pointer compilation flag under GNU/Linux
— drop -static-libstdc++ flag under Windows
— no longer link to -lstdc++ under GNU/Linux and macOS
— gracefully fail when 'mexext' value is unknown
2021-05-24 18:21:23 +02:00
a2c96a9d79
Allow empty quoted strings
...
Can be useful when resetting a filename option to an empty value, e.g. for the
mh_initialize_from_previous_mcmc_* options.
Ref. dynare!1841
2021-05-11 18:50:18 +02:00
73e07b78b8
Add mh_initialize_from_previous_mcmc option family to estimation command
...
Ref. dynare!1841
2021-05-11 18:12:54 +02:00
be7ff255f6
The fast_realtime option now accepts a vector of integers
...
Ref. dynare!1842
2021-05-10 15:41:36 +02:00
80e1090257
Add no_init_estimation_check_first_obs option to estimation
...
Ref. dynare!1839
2021-05-07 17:41:20 +02:00
337b6f469c
New parallel_use_psexec command-line option
...
Ref. dynare!1843
2021-05-06 16:25:46 +02:00
1d188a7444
Add comment
...
[skip ci]
2021-05-04 17:21:47 +02:00
3c4d647cc3
Forbid mfs > 0 in a stochastic setup (or even when using check)
...
The result would be wrong.
Ref. dynare#1726
2021-05-04 17:11:09 +02:00
7765efe95b
Windows: fix detection of location of MinGW compiler
2021-04-30 15:03:53 +02:00
12b61e4817
Bugfix in DataTree::writeToFileIfModified()
...
If the old file and the new contents shared a common prefix, then the output in
the generated file would be truncated at the start.
Ref. DynareJulia/Dynare.jl#1
2021-04-26 14:10:47 +02:00
9773bafc5b
Julia: write definition of get_power_deriv inside generated static/dynamic files
2021-04-26 10:52:18 +00:00
dca294de3a
Julia: refactor the auxiliary variable/series files
...
— rename the files with camel case
— encapsulate the functions with modules
— change the signature of function <MODFILE>_dynamic_set_series!, by removing
the output argument and appending the exclamation mark, since this function
modifies one of its arguments, and for symmetry with the static version
Ref. DynareJulia/Dynare.jl#1
2021-04-23 17:57:33 +02:00
4354d8a578
Julia: do not overwrite steady state / auxiliary variables/series files if contents has not changed
...
Ref. DynareJulia/Dynare.jl#1
2021-04-23 17:56:54 +02:00
4cfc7a563b
Julia: do not overwrite Static/Dynamic files if contents has not changed
...
Ref. DynareJulia/Dynare.jl#1
2021-04-22 18:03:47 +02:00
81b1d02748
Julia: fix typo in StatsFuns package name
...
Ref. DynareJulia/Dynare.jl#1
2021-04-22 15:48:25 +02:00
0213c9be73
Merge branch 'ramsey_k_order' into 'master'
...
Allows to write M_.NNZDerivatives_objective in the driver.m file
See merge request Dynare/preprocessor!36
2021-04-22 13:28:18 +00:00
22fbfb0fba
Allows to write M_.NNZDerivatives_objective in the driver.m file
...
Writes the number of non-zero derivatives of the planner objective function NNZDerivatives_objective in the driver.m file. Useful for the k-order approximations of welfare.
2021-04-22 15:26:25 +02:00
d511e3ec3a
add deterministic_trends statement (only used by Dynare Julia)
2021-04-22 15:19:21 +02:00
910649dd28
Fix treatment of “series” option of “initval_file”
...
This option was treated as a numeric value: it was a hack to ensure that the
name of the dseries object was printed without quotes in the driver file.
However, that would make the JSON output synctically incorrect.
This commits removes the hack, and treats the “series” option as a string
value (in the same way as for the “data” command). A modification of the MATLAB
code will thus be needed to accomodate the change.
2021-04-20 10:39:14 +02:00
1f17528383
Julia: update dependencies of Static/Dynamic modules
...
Ref. DynareJulia/Dynare.jl#1
2021-04-19 17:02:46 +02:00
c2d47fb48e
Julia: remove driver file
...
The Julia code now uses the JSON output for retrieving the information that
previously was in the driver file.
2021-04-19 17:02:41 +02:00
7b3df21f11
DataTree::operator=(): fix crash when symbol ID ordering of model local variables does not correspond to the recursive ordering in the model block
...
Ref. dynare#1782
2021-04-16 17:34:13 +02:00
c2009e2012
fix set_auxiliary_variables.jl
2021-04-01 17:05:13 +02:00
6662d8014e
fix writeJsonOutput for load_params_and_steady_state
2021-04-01 17:05:13 +02:00
fd979f625a
add maximum lead and lag to modfile.json
2021-04-01 17:05:13 +02:00
c9dd241f5b
add varobs statement to JSON output
2021-04-01 17:05:12 +02:00
50a2e7c44f
Add escape codes in NativeStatement and VerbatimStatement writeJsonOutput() method
2021-04-01 17:05:10 +02:00
29b3b744ef
Remove spurious space produced when parsing key/value pairs
...
Introduced in e46168db66
2021-02-26 12:28:06 +01:00
Stéphane Adjemian (Charybdis)
f0d67b01dd
Honour bi-annual dates in mod file.
2021-02-26 09:30:02 +01:00
5581052451
Add particle_filter_options option to estimation
...
Ref. #59
2021-02-23 16:54:36 +01:00
e46168db66
Simplify parsing of lists of key/value pairs options
2021-02-23 16:49:11 +01:00
54365b16ed
Remove unfinished “occbin” option to perfect_foresight_solver and extended_path
...
Partially reverts f914837aff567156aad4
2021-02-22 16:22:17 +01:00
825b9ee8ef
Fix bugs introduced in d3e115194b
...
Ref. #65
2021-02-19 20:29:18 +01:00
4798a5e2bc
Cosmetic change
2021-02-19 20:09:48 +01:00
ca4b1ba4ed
Remove time subscript in LaTeX output of model-local variables
...
See https://forum.dynare.org/t/tex-output-unwanted-subscript-t/16072
2021-02-19 17:45:58 +01:00
d3e115194b
Fix output write_latex_original_model for model-local variables
...
Those variables were substituted out by early substitution methods (which are
also called on the original model, for an unknown reason).
Closes : #65
2021-02-19 17:45:02 +01:00
071a46a798
Add interface to set zero_moments_tolerance
...
Useful in identification and method_of_moments toolbox to set the tolerance which replaces very small numbers by hard zeros in unconditional second moments (particularly in pruned_state_space_system.m and disp_moments.m)
2021-02-16 16:29:57 +01:00
2781e3e6f6
A few modernizations using clang-tidy
2021-02-10 16:48:46 +01:00
9274675efb
Allow booleans in optimization options list
...
Closes : #63
2021-02-10 16:18:58 +01:00
f7cd6dbf0a
🐛 save oo_.time also in _results.mat
2021-02-04 10:48:59 +01:00
702cf62ee3
Declare some methods as const
2021-02-03 18:09:11 +01:00
234fcd1d2b
Remove obsolete stuff from bytecode
2021-02-01 14:46:08 +01:00
754a1a464d
Use default member init
...
Performed using modernize-use-default-member-init from clang-tidy.
https://clang.llvm.org/extra/clang-tidy/checks/modernize-use-default-member-init.html
2021-02-01 12:44:24 +01:00
4ee89d6e70
Use more auto keyword
...
Performed using modernize-use-auto from clang-tidy.
https://clang.llvm.org/extra/clang-tidy/checks/modernize-use-auto.html
2021-02-01 12:42:59 +01:00
07391e0f9f
Remove unused kludge
2021-02-01 12:07:33 +01:00
89e2daf7c7
C++17 modernization: convert the remaining enums to enum classes
2021-02-01 12:05:06 +01:00
c79ae5d4bd
Save 'Total computing time' into oo_.time
2021-01-28 11:28:41 +01:00
ad22e62874
filter_initial_state: Port lost preprocessor commits from https://git.dynare.org/JohannesPfeifer/dynare/-/commits/filter_initial_state
2021-01-26 16:03:04 +01:00
7468b4d8d7
Remove more stuff related to unfinished decomposition in linear and nonlinear blocks
...
Ref. #39 , dynare!1626
2021-01-25 18:13:35 +01:00
af9bd8f109
Remove the unfinished decomposition in linear and nonlinear blocks
...
Also remove the det_cond_forecast interface, that was related to it.
Ref. dynare!1626
Closes : #39
2021-01-25 18:03:37 +01:00
1e5de7d71e
Fix typo
...
Introduced in 4c14dc3fbf
2021-01-22 14:20:52 +01:00
4c14dc3fbf
Macro processor: gracefully fail when argument to @#ifdef/@#ifndef is not a variable name
...
Ref. dynare#1771
2021-01-22 12:20:13 +01:00
ec66768a63
Forbid exogenous deterministic variables with a lead or lag
...
Ref. dynare#1765
2021-01-22 11:59:44 +01:00
ab656f2d84
macOS/use_dll: upgrade to gcc-10 compiler
2021-01-13 15:42:09 +01:00
a0d12310cb
macOS/use_dll: drop -mmacosx-version-min=10.9 flag
...
The MEX files are not intended for distribution to other systems.
2021-01-13 15:42:09 +01:00
c91178e3e3
macOS/use_dll: no longer pass -arch x86_64, in preparation for ARM support
2021-01-13 15:42:09 +01:00
Stéphane Adjemian (Charybdis)
7a8446e836
Added call to a routine filling the fields required by some nonlinear solvers.
...
At some point this task might be delegated to the preprocessor.
(cherry picked from commit ca230d0902eadeb89ab705cf66790ffa2f135192)
2021-01-12 19:50:37 +01:00
a87b2753ed
MoM: Add option analytic_jacobian
2021-01-08 18:57:06 +01:00
8e03f17350
Cleanup the “output” option
...
— output Julia files as soon as “language=julia” is passed, independently of
the value of the “output” option
— drop the “dynamic” and “first” values of the “output” option, since they
actually do nothing
— obey the “output” option even in the deterministic case
Ref. dynare#1600
2021-01-08 18:51:48 +01:00
a9daec1c07
Rename preprocessor binary to dynare-preprocessor
...
Ref. dynare#1728
2021-01-06 17:08:42 +01:00
cd443a6083
Revert to accepting the “nolog” option, but do nothing about it
...
This is necessary to avoid an “unknown option” error.
Ref. dynare#1743
2021-01-06 15:50:30 +01:00
78874a01d3
No longer deal with the log file at the preprocessor level
...
Ref. dynare#1743
2021-01-06 14:19:33 +01:00
1def52a088
Add interface to set schur_vec_tol
2020-12-17 19:55:20 +01:00
91db87b1cf
Forbid some basenames, since they will cause trouble
...
Closes: preprocessor#62
2020-12-16 18:21:55 +01:00
84d79226cd
Lexer: no longer associate a semantic value to GMM and SMM tokens
...
This was forgotten from commit 78f6d48601
2020-12-14 19:10:29 +01:00
78f6d48601
GMM: introduce option analytic_standard_errors
2020-12-12 09:36:40 +01:00
7baeda8b8d
Occbin: add equation number in M_.occbin.constraint structure
...
Ref. #569
2020-12-10 15:57:47 +01:00
f67e9c044b
JSON output: fix “variable_mapping” for Ramsey models
...
Ref. dynare#1755
2020-12-09 17:32:29 +01:00
bf1da3c6fb
JSON output: fix logic in “variable_mapping” that could lead to invalid JSON formatting
...
Ref. dynare#1755
2020-12-09 16:47:55 +01:00
ec779235ea
Macro-processor: move environment out of node classes
...
This completes the separation of parse trees from their evaluation.
Closes : #32
2020-12-07 17:29:05 +01:00
638b4c7ac5
Ensure that all statements produce an output in JSON mode
2020-11-26 16:25:22 +01:00
a99ca9f450
Remove obsolete “var_estimation” statement
2020-11-26 16:13:52 +01:00
1b40327639
Remove obsolete “restrictions” block
2020-11-26 16:13:49 +01:00
c697a63119
Fix typo
2020-11-25 18:08:49 +01:00
9292b0692e
Rename DetCondForecast to DetCondForecastStatement, for naming consistency
2020-11-25 17:40:00 +01:00
4b8290faa5
Do not crash when unknown symbol is passed to “irf_shocks” option
...
Closes : #57
2020-11-18 17:47:37 +01:00
b1df9ecc38
Parallel config file: switch back to DynarePath field (instead of ProgramPath)
...
This is a partial revert of f022f666b7
2020-11-13 15:35:01 +01:00
cfb41d291c
Substitute out model-local variables early in the model transform pass
...
Incidentally, this makes it possible to use model-local variables with either
the “block” or “bytecode” option. (Ref: dynare#1243)
2020-11-10 18:07:57 +01:00
1e419a48e4
Remove unused function
2020-11-10 18:02:06 +01:00
9bca65c7e5
Deterministic models: replace exogenous with lead/lags by auxiliary variables
...
This brings those models in line with stochastic models.
However note that transformation is still not exactly the same on stochastic
and deterministic models, because there is no need to take into account the
Jensen inequality on the latter. In deterministic models, there is a one-to-one
mapping between exogenous with lead/lags and auxiliaries, while in stochastic
models, an auxiliary endo may correspond to a more complex nonlinear expression.
2020-11-09 16:28:32 +01:00
5d564eed07
Fix the interpretation of @#elseif clauses within an @#ifdef or @#ifndef
2020-11-02 12:47:33 +01:00
df8d8ef116
Fix logic for @#ifndef
...
Ref. dynare#1747
2020-11-02 12:27:27 +01:00
Stéphane Adjemian (Charybdis)
abf5a6553b
Removed useless subfield (only julia output).
...
TODO: Check that we can do the same in the matlab version…
2020-10-21 18:16:07 +02:00
Stéphane Adjemian (Charybdis)
b8f7b5846b
Fixed flags for temporaries (julia output).
2020-10-21 16:37:13 +02:00
56581b1dd4
PAC model: make detection of non-optimizing part more robust
...
Introduce a new method for decomposing a product of factors, so that we can
identify expressions of the form (1-optim_share)*A*B.
Also enforce that the optim_share parameter be in a factor of the form
1-optim_share (previously it would accept any expression containing the
parameter).
Note that this fix does not yet allow to actually write non-optimizing parts of
the form (1-optim_share)*A*B, since at a later point the preprocessor imposes
that this part be a linear combination of variables (but in the future we could
think of expanding the A*B product into a linear combination if, for example, A
is a paramater or a constant and B is a linear combination).
Closes : #50
2020-10-19 18:32:30 +02:00
db3ef925a7
Fix handling of diffLead in SymbolTable::getOrigLeadLagForDiffAuxVar()
2020-10-16 16:31:55 +02:00
d2cbe2a19e
PAC model: fixes to M_.pac.MODEL.ar (information about autoregressive part)
...
– Fix order of items in this structure. Previously, items were ordered
according to the declaration order of parameters. Now, items are order
according to lag order (first lag appears first)
– Gracefully handle the case where there is no autoregressive part
(Closes : #52 )
2020-10-16 16:31:51 +02:00
740ea833f6
Block decomposition: bugfix for solve one boundary blocks with mfs>0
...
Ref. dynare#1726
2020-10-07 18:55:19 +02:00
e736b9234f
Minor simplification
2020-10-07 17:15:53 +02:00
4560639eb4
Block decomposition: fix evaluation of recursive variables in solved blocks
...
Ref. dynare#1726
2020-10-06 18:30:34 +02:00
5979885714
Block decomposition, chain rule derivation: code improvement
2020-10-02 18:31:55 +02:00
d347a92f14
Drop ExprNode::containsExogenous()
...
Replace it by calls to ExprNode::collectVariables() at the single place where
it was used.
2020-09-29 18:41:31 +02:00
063207e71f
ExprNode::collectEndogenous() is not a virtual method
2020-09-29 18:41:31 +02:00
2424b7358c
Remove unused ExprNode::collectExogenous()
2020-09-29 18:41:31 +02:00
b0bd031a9f
Remove unused ExprNode::containsEndogenous()
2020-09-29 18:41:31 +02:00
7f05489639
Simplify ExprNode::findConstantEquations()
2020-09-29 18:41:31 +02:00
a377fc8398
Fix various potential bugs with model local variables
...
Many recursive function on ExprNodes were not correctly recursing into the
definition of model local variables.
2020-09-29 18:41:17 +02:00
e4687bb976
Ensure that negative constants are never marked as temporary terms
2020-09-23 17:28:53 +02:00
571ea3eb8a
SymbolTable: add some const qualifiers
2020-09-23 17:28:04 +02:00
8f31e443d6
SymbolTable::getTypeSpecificID() now throws an exception instead of returning -1 when symbol does not have a type-specific ID
...
This can potentially avoid some undetected bugs.
2020-09-23 17:09:36 +02:00
7ba77751c3
simplify external function C output
2020-09-18 17:05:54 +02:00
98d01cbbb6
Provide a meaningful error message when 'use_dll' option is used without 'matlabroot'
2020-09-18 15:00:47 +02:00
7e1563fdbb
Allow 'use_dll' option to be specified on the command line
...
Ref. Enteprise/dynare#4
2020-09-18 15:00:00 +02:00
Stéphane Adjemian (Charybdis)
0f268ce70c
Fixed logic.
...
Value passed to steady_state_growth was not acknowledge.
2020-09-18 11:11:50 +02:00
feee0ffa12
PAC model: better error message when pac_expectation refers to a non-existent pac_model
...
Closes : #53
2020-09-10 14:24:20 +02:00
de65e74c8f
use_dll: add the “restrict” C99 keyword to all pointers to double in computation functions
...
This greatly facilitates the job of the compiler during the optimization pass,
since we promise that the various pointers do not overlap each other. It may
now be possible to reenable some of the optimization flags that were disabled
without sacrificing compilation time, but this needs more investigation.
For the gory details, see:
https://en.cppreference.com/w/c/language/restrict
https://cellperformance.beyond3d.com/articles/2006/05/demystifying-the-restrict-keyword.html
2020-09-03 18:09:06 +02:00
5a9d1c4d87
Merge branch 'MoM' into 'master'
...
Remove matched_moments_ as input of method_of_moments
See merge request Dynare/preprocessor!27
2020-09-01 13:13:26 +00:00
c6ad212c8a
Merge branch 'typo' into 'master'
...
Fix typo
See merge request Dynare/preprocessor!26
2020-09-01 12:15:44 +00:00
9d9a4bd4d1
Remove matched_moments_ as input of method_of_moments
...
Now a subfield of M_
2020-08-05 13:46:09 +02:00
d714cac8e1
EquationTags.hh: fix typo
2020-08-04 16:21:50 +02:00
09054e11dc
Add new “matched_moments” block
...
Closes : #51
2020-07-28 18:30:24 +02:00
ecf6d67e1c
PAC/MCE model: fix timing of target in PAC expectation term
...
Also be more strict on the form of the target (must now be X(-1) or log(X(-1))
where X is *not* an auxiliary variable).
By the way, improve some comments in SymbolTable.
2020-07-24 18:26:16 +02:00
3b3589b958
PAC model: rewrite the algorithm for detecting the target, to make it more robust
...
Closes : #54
2020-07-23 17:19:16 +02:00
7bca47385d
Fixes to include_eqs/exclude_eqs
...
— fix segfault due to bad handling of iterators while erasing elements of a set
— fix logic for excluding endogenous variable when it is present on the LHS
2020-07-23 11:06:57 +02:00
86ac7d5bbc
PAC/MCE: improve informative message about # of aux. vars/eqs by adding the model name
...
This makes the output more useful when there are multiple PAC models.
2020-07-22 16:09:12 +02:00
2b25ab3e85
Fix handling of multiple PAC models
2020-07-22 12:21:05 +02:00
f2f6097777
Merge branch 'tune_jscale' into 'master'
...
Fix faulty logic in checks for mh_tune_jscale
Closes #46
See merge request Dynare/preprocessor!25
2020-07-21 08:59:20 +00:00
eb464665ac
Fix faulty logic in checks
2020-07-21 10:48:45 +02:00
5cc9f216ca
Make preprocessor recognize default order=2 of stoch_simul
2020-07-21 10:40:08 +02:00
d4e20c7ec4
Merge branch 'moment_estimation' into 'master'
...
Method of moments: Preprocessor preparations
See merge request Dynare/preprocessor!23
2020-07-16 08:56:03 +00:00
Stéphane Adjemian (Charybdis)
03f53c1b0d
Allow empty shocks(overwrite) block.
2020-07-15 12:33:02 +02:00
16f9168fda
PAC MCE: fix incorrect detection of the target variable
...
The detection of the target EC variable to be used when constructing the
forward-looking expectation variable is rather fragile.
When the PAC model is written with an (non-)optimizing share of agents,
restrict the identification of the target variable to the optimizing
expression, to minimize the risk of wrong identification.
By the way, add a few comments, and a small simplification.
2020-07-13 17:52:42 +02:00
98ddbb61ba
Method of moments: Preprocessor preparations
2020-07-10 11:18:57 +02:00
Stéphane Adjemian (Charybdis)
51b3d5c68f
Do not print aux. var. for PAC expectation in dynamic_set_auxiliary_series routine.
...
It is useless in the backward looking version, and it cannot be
evaluated in the MCE case (because of the forward looking recursion).
2020-07-08 22:51:35 +02:00
3a9b0fc9ce
PAC MCE expectations: add corresponding auxiliary variables in dynamic_set_auxiliary_series.m and set_auxiliary_variables.m
2020-07-08 22:49:51 +02:00
4b3f010bc6
Simplify DynamicModel::writeAuxVarRecursiveDefinitions()
2020-07-08 18:48:45 +02:00
5b80a4db59
Unary ops substitution: bugfix, actually restrict the substitution to the chosen equations
...
By default, the preprocessor is supposed to only do the “unary ops”
transformation in the equations of VAR/PAC/trend component models.
However, the implementation was slightly different so far. It would detect
candidates to this transformation in the chosen equations, but it would then
perform the substitution in *all* equations.
This could lead for crashes, for example if the chosen equation contains
log(X(-1)), but another (non-chosen) equation has log(X(-2)). Then this latter
expression, even though it belongs to the same lag-equivalence class, is not
properly handled, causing a segfault.
Also do a few related cosmetic changes.
2020-07-08 15:13:14 +02:00
59bda04d8b
Block decomposition + use_dll: fix bug in timing of exogenous variables
2020-06-25 15:19:49 +02:00
c4351166a9
use_dll: remove nb_row_x from static file, it is useless
2020-06-23 18:19:27 +02:00
db8c5439f7
use_dll: higher order derivatives are now returned as sparse matrices by static/dynamic files
...
Previously they were returned as 3-column matrices. But this was inconsistent
with the M-file mode.
2020-06-23 17:50:50 +02:00
95bdfcee8c
use_dll: better safeguards on the number of input/output argument to static/dynamic MEX files
2020-06-23 16:30:21 +02:00
633b89db64
use_dll: remove hack for Microsoft Visual C++
...
We don’t support that compiler, and in any case it should support C99.
2020-06-23 16:00:55 +02:00
fd13c03401
use_dll: use fmax() and fmin() from C99 to compute max and min operators
2020-06-23 15:59:35 +02:00
ad5e196d30
Block decomposition now compatible with 'use_dll' option
2020-06-23 15:13:04 +02:00
479c2c029f
Block decomposition: move core of the routine for writing per-block files in separate function
...
This is a preparatory step to allow use_dll with block decomposition.
2020-06-22 14:18:21 +02:00
7641c2f7ee
Cosmetics
2020-06-22 11:33:16 +02:00
f098329c74
Block decomposition: static and dynamic files are now pure wrappers around the per-block files
2020-06-19 18:45:42 +02:00
f4d2ce70da
Block decomposition: move logic for Jacobians in stochastic mode outside the dynamic file
2020-06-19 17:44:05 +02:00
92aff91066
Block decomposition: in per-block files, construct sparse Jacobians more efficiently
...
Use 3-column format before calling sparse().
Ensure that the 3-column matrix is constructed in column-major order: data
locality will greatly improve performance once we implement use_dll.
2020-06-19 16:42:32 +02:00
93c18d514b
Block decomposition: in per-block dynamic files, fix nonzero elements for deterministic Jacobians
2020-06-19 15:26:19 +02:00
8ec43b1fb2
Block decomposition: small simplification in per-block dynamic files
2020-06-19 12:39:54 +02:00
dc28aaa5e9
Block decomposition: fix bug in reduction of block decomposition
...
Some evaluated blocks could be incorrectly merged, leading to wrong
results (e.g. blocks incorrectly marked as evaluable forward).
Bug introduced in 39407083be
2020-06-17 19:42:13 +02:00
79763911b2
Block decomposition: in the dynamic file, y and T are now vectors as in non-block mode
2020-06-17 16:49:12 +02:00
91d5be3b8c
Block decomposition: fix value of M_.block_structure.dyn_tmp_nbr
2020-06-17 16:04:49 +02:00
257b399a94
Block decomposition: in dynamic files, rename “jacobian_eval” to “stochastic_mode”
...
This is much more explicit.
2020-06-16 15:54:51 +02:00
379be6ccef
Block decomposition: for “solve two-boundaries” blocks, move the iteration loop outside the dynamic file
2020-06-16 15:34:45 +02:00
85b900363d
Block decomposition: for “solve two boundaries” block, no longer compute “b” (-residuals+g1*y) in the dynamic function
2020-06-16 13:49:23 +02:00
8496979375
Block decomposition: remove useless “y” output for static/dynamic files of solved blocks
2020-06-16 10:38:16 +02:00
ad583adcd4
Block decomposition: for evaluate blocks, move the iteration loop outside the dynamic file
2020-06-15 18:13:17 +02:00
f022c0e16e
Block decomposition: remove duplicate “periods” arg. to dynamic file of complete two-boundaries blocks
2020-06-15 16:54:11 +02:00
d530ccd012
Misc cosmetic improvements
2020-06-05 17:12:15 +02:00
b402e35683
In the driver file, provide mapping between model local variables and indices in the temporary terms vector
...
The structures are M_.model_local_variables_{dynamic,static}_tt_idxs.
Ref. dynare#1722
2020-06-05 17:12:15 +02:00
90b9b6a1e4
JSON output: write model local variables in parsed model (modfile.json)
...
Ref. dynare#1723
2020-06-05 16:10:38 +02:00
7b9495bb2f
JSON output: when writing the model after computing pass, use TEF terms for (not derived) external functions
2020-06-05 16:10:38 +02:00
f9ab44da56
JSON output: do not append two underscores after model local variable names in their definition
...
This is inconsistent with the way they are printed in equations (without
underscores).
The practice of appending underscores only makes sense in a MATLAB workspace context.
2020-06-05 16:10:37 +02:00
931935a942
Bug fix: allow external functions in model local variables
2020-06-05 14:50:06 +02:00
050f1cdd74
Block decomposition: remove unused 2nd and 3rd derivatives in output of dynamic functions
2020-06-03 16:59:04 +02:00
f0c8a14551
Block decomposition: no longer compute deterministic simulation from dynamic.m file
2020-06-02 17:42:00 +02:00
2caad3ee84
Block decomposition: simplify DynamicModel::writeDynamicBlockMFile()
2020-05-29 16:48:38 +02:00
7b11898689
use_dll: merge the two generated C source files into a single one
...
Also, do various other cosmetic/simplification changes.
2020-05-29 16:12:01 +02:00
3c5d731500
Fix compilation with GCC 10
2020-05-28 14:21:58 +02:00
3ca1914e03
Block decomposition: fix simulation of blocks of type "evaluate backward" (without bytecode)
...
Ref. dynare#1727
2020-05-26 17:00:21 +02:00
faa6666abe
Block decomposition: no longer use global variables for temporary terms in the dynamic M-file
...
Rather use a single vector as in non-block mode.
By the way, change the order of output arguments in static functions, to be
closer to the dynamic ones.
2020-05-26 16:21:11 +02:00
04b7d4386d
Block decomposition: no longer use global variables for temporary terms in the static M-file
...
Rather use a single vector as in non-block mode.
2020-05-25 18:35:36 +02:00
a5abbae46a
use_dll: remove workaround for char16_t type
...
Should no longer be needed now that the minimal MATLAB version is R2014a.
2020-05-25 16:06:17 +02:00
b9216be277
Block decomposition: fix incorrect check with model local variables
...
We normally never arrive here, since MLV are not compatible with 'block'
option.
2020-05-25 15:42:16 +02:00
3033417f5a
Block decomposition: remove ExprNodeOutputType::matlabDynamicSparseSteadyStateOperator
...
It’s not really needed, since it does the same as
ExprNodeOutputType::matlabDynamicSteadyStateOperator.
2020-05-25 15:35:55 +02:00
7b1a77ca47
Block decomposition: use logical values for some inputs to solve_{one_boundary,two_boundaries}
2020-05-25 15:12:43 +02:00
a6d9ba6e55
Rename EquationType::evaluate_s to evaluateRenormalized, for clarity
2020-05-20 11:49:32 +02:00
c6a17913b4
Rename {Static,Dynamic}Model::writeOutput() to ::writeDriverOutput() for clarity
2020-05-20 11:44:40 +02:00
becc433d56
Rename byte_code to bytecode for consistency with user option name
2020-05-20 11:39:59 +02:00
c579a45b08
Remove helper now unused because of 1d838e96ff
2020-05-20 11:35:44 +02:00
96657b4974
Block decomposition: refactor code that writes the block structure to the driver file
2020-05-20 11:35:14 +02:00
4b28f1fe9c
Bytecode: remove unused data structure
2020-05-20 10:31:31 +02:00
2944484442
Bytecode: rename output methods for better consistency
2020-05-19 17:45:33 +02:00
1d838e96ff
Block decomposition: simplify routines for writing output dynamic/static M files
2020-05-19 17:33:58 +02:00
d05ffde63e
Merge branch 'hif' into 'master'
...
histval_file initval_file refactoring
See merge request Dynare/preprocessor!20
2020-05-14 14:36:36 +00:00
ed4c36a267
add first_simulation_period to the lexer
...
(cherry picked from commit c8575e11d2f81ad606ab50849c52127653699a93)
2020-05-14 14:36:35 +00:00
a2bea00fee
Block decomposition: another fix related to temporary terms refactoring
...
Temporary terms need to be computed per equation (as was done previously), and
not simply per block.
It’s necessary to track temporary terms per equation, because some equations
are evaluated instead of solved, and an equation E1 may depend on the value of
an endogenous Y computed by a previously evaluated equation E2; in this case,
if some temporary term TT of equation E2 contains Y, then TT needs to be
computed after E1, but before E2.
2020-05-13 16:58:19 +02:00
c9388da649
Bytecode: still another fix for temporary terms
2020-05-13 15:16:38 +02:00
2e6bc7ecaa
Bytecode: fix handling of temporary terms
...
They were broken by 49036f166d
2020-05-13 12:53:47 +02:00
f8796c67c5
ExprNode::computeTemporaryTerms: minor simplifications, for symmetry with blocks code
2020-05-12 18:39:15 +02:00
49036f166d
Block decomposition: refactor code for computing temporary terms
...
In particular, in dynamic models, temporary terms are now computed for
derivatives w.r.t. exogenous, and also w.r.t. endogenous variables that do not
belong to the block.
2020-05-12 18:39:13 +02:00
ba214a100d
Block decomposition: bugfix, cutoff option was ignored
...
Introduced in c699d57308
2020-05-07 18:41:01 +02:00
928aa977b6
Block decomposition: replace expr_t by BinaryOpNode * in a couple of places
2020-05-07 15:24:35 +02:00
c699d57308
Block decomposition: no longer compute static Jacobian, it’s not used
...
Also remove a message about elements below the cutoff that was no longer
correct (elements below the cutoff have no impact on the incidence matrix
outside of normalization).
2020-05-06 18:11:11 +02:00
a1b37288f9
Block decomposition: do not write unnecessary zeros in dynamic Jacobian
...
Also do cosmetic change in the corresponding code for static Jacobian.
2020-05-06 14:22:04 +02:00
9f361e9e37
Block decomposition: simplify code for ordering dynamic Jacobian columns
2020-05-06 14:21:24 +02:00
061245e50d
Block decomposition: simplify routines that write static model in MATLAB form
...
By the way, this probably fixes a bug in the presence of external
functions (so-called TEF terms were not properly repeated in each per-block
static file).
Also remove debugging output in the M-file.
2020-05-05 18:59:28 +02:00
38ccd5e0cf
Block decomposition: remove unneeded code for bytecode
2020-05-05 17:15:08 +02:00
5f22f00436
PAC models: handle PAC data in DynamicModel copy constructor and assignment operator
2020-05-05 16:34:51 +02:00
4f453320d1
Block decomposition: remove unneeded code for exogenous
2020-05-05 16:17:22 +02:00
cb61ee8506
Block decomposition: make structures storing derivatives w.r.t. other endo/exo symmetric with those for endos belonging to the block
2020-04-30 18:24:24 +02:00
b1a739c56a
DynamicModel: add missing members in copy constructor/assignment operator
2020-04-30 14:57:50 +02:00
a8c2defe26
Block decomposition: compute simulation types into separate function; prologue and epilogue are no longer class members
...
By the way, remove the BlockType stuff which was purely informative (and it’s
not worth carrying over prologue and epilogue information just for that).
2020-04-30 12:48:16 +02:00
26fc71ffa7
Block decomposition: store max lead/lag of all variables in the “blocks” structure
...
By the way, regroup all code that computes the dynamic structure of a block
into a single function.
2020-04-30 12:03:14 +02:00
7f57821401
Now compute blocks[].first_equation from ModelTree::computeBlockDecompositionAndFeedbackVariablesForEachBlock
...
Also include various cosmetic changes.
2020-04-29 18:48:42 +02:00
1886746581
VariableDependencyGraph::reorderRecursiveVariables now returns indices from vertex_index
...
This is for consistency with other graph algorithms. Previously it would return
an index from vertex_index1.
See also 1d92adacf4
2020-04-29 18:46:15 +02:00
24db5e6033
Minor simplifications in graph algorithms
2020-04-29 18:12:22 +02:00
4948adf2cd
Block decomposition: remove unused data structure
2020-04-29 17:47:48 +02:00
f39c687d85
Block decomposition: store mapping between endos/eqs and their block
...
Previously this computation was done at several places.
2020-04-29 17:42:46 +02:00
39407083be
Block decomposition: ensure that the cutoff option only impacts normalization
...
Previously, the cutoff option would also impact the block decomposition itself,
since it would had an influence on the incidence matrix used for computing the
blocks and their derivatives.
The problem is that, in the general case, it’s quite possible that an element
of the numerical Jacobian be zero at the evaluation point, while being quite
different from zero along the simulation path. A typical example is an
expression of the type x*y, where y is an endogenous and x is an exogenous not
present in the initval block (and hence initialized to zero).
2020-04-29 12:41:39 +02:00
470cb5fcb0
Block decomposition: remove mostly unused structure ModelTree::derivative_endo
...
It has been superseded by ModelTree::blocks_derivatives.
By the way, fix the initial number of non-zero elements in sparse Jacobian.
Also avoid computing suboptimal temporary terms.
2020-04-28 14:12:29 +02:00
2c69f1f31e
Cosmetic changes to ModelTree::determineLinearBlocks()
2020-04-28 12:35:18 +02:00
2e136cae56
Block decomposition: simplify computation of block derivatives
...
— use a std::map for storing block derivatives
— remove redundant ModelTree::first_chain_rule_derivatives structure
— remove unused codepaths in StaticModel
— DynamicModel: simplify code that determines the type of derivatives in a
block. We now use a slightly different categorization.
— by the way, fix the max lead/lag information for blocks that are obtained via
merging. A workaround was previously implemented in
DynamicModel::get_Derivative(), but it is no longer needed with this fix.
2020-04-28 12:26:51 +02:00
5bd41f8599
Improve comments
2020-04-24 19:06:30 +02:00
45b260cf20
VariableNode::getChainRuleDerivative: do not cache values in ExprNode::derivatives
...
This field is used for standard derivatives. Using it also for chain rule
derivatives can only lead to wrong results.
2020-04-24 19:06:30 +02:00
29f28be772
Further simplify ModelTree::computeBlockDecompositionAndFeedbackVariablesForEachBlock()
2020-04-24 11:36:05 +02:00
5ee45f5bd4
Block decomposition: remove helpers that are no longer needed since data structure are cleaner
...
Add a new helper for getting the number of recursive variables in a block.
2020-04-23 16:31:57 +02:00
2154227990
Block decomposition, bugfix: M_.block_structure_stat.block().mfs was wrong
2020-04-23 16:03:38 +02:00
e48e761b94
Block decomposition: various minor cleanups
2020-04-23 14:41:18 +02:00
8b6e42559d
fix typo
...
[skip ci]
2020-04-21 13:08:27 -04:00
118ceab85b
Block decomposition: reorganize data structures storing block information
2020-04-21 18:25:33 +02:00
8eafd9ab4f
Block decomposition: various refactorings
2020-04-17 19:23:52 +02:00
1d92adacf4
Block decomposition: fix bug with mfs > 0 introduced in 5431451db3
2020-04-17 18:19:07 +02:00
a741799f90
Block decomposition: factorize various helper functions
...
Should have been done in 76c2e87c3b
2020-04-17 16:04:19 +02:00
7327fb9f17
Block decomposition: give more explicit names to ModelTree::{inv_,}{equation,variable}_reordered
2020-04-17 15:09:22 +02:00
Sébastien Villemot
MichelJuillard
Johannes Pfeifer
Stéphane Adjemian (Charybdis)
NormannR
Willi Mutschler
Houtan Bastani