8c5d6f5eae
libkordersim: ensure strict Fortran standard compliance
...
The “order” variable was used before being typed. This is apparently accepted
by the GNU Fortran dialect.
2023-09-16 14:25:41 +02:00
9e7e069aa8
Fortran: use standard-compliant ieee_is_nan() instead of GNU extension isnan()
2023-09-16 14:25:40 +02:00
bd9943a695
Bytecode: fix memory leak in Evaluate class
...
The newly-created FBEGINBLOCK_ and FCALL_ instances were freed using a
base-class pointer. But the latter does not have a virtual destructor.
Those class instances are now stored by value in containers, so that the
destructor of the derived class is used.
2023-09-01 14:52:57 +02:00
b4f260aa01
folded_to_unfolded_dr MEX: remove unused variable
2023-07-27 17:29:32 +02:00
b1db895dee
block_trust_region MEX: remove unused parameter
2023-07-27 17:29:31 +02:00
0442fe785e
libkordersim: only BLAS is used, not LAPACK
2023-07-27 09:44:34 +02:00
33719e69f2
Build system: libkorder no longer depends on MatIO
2023-07-26 16:30:51 +02:00
9ff1d8556c
libkorder: remove mention of Dynare++ in the journal
...
By the way, use PACKAGE_VERSION macro (instead of VERSION), for consistency
with the preprocessor.
2023-07-26 16:30:51 +02:00
710589eb5b
Adds pruning in k_order_simul
...
See issue #1643 about beyond-third-order pruning
2023-07-20 06:58:35 +01:00
5b8b63924d
Remove unneeded braces
2023-07-19 18:03:18 +02:00
661e5c7e9f
C++17 modernization: initialization in “if”-statement
2023-07-19 18:03:03 +02:00
23dbb2b4b9
Fixes local_state_space_iteration_3 and adds a test comparing the various pruning routines used in dynare up to order 3.
...
The considered pruning routines can be found in
- `simult_.m`
- `local_state_space_iteration_2` MEX
- `local_state_space_iteration_3` MEX
- `pruned_state_space_system.m`
2023-07-04 19:15:43 +02:00
8b945dfd64
libkorder: use infinity instead of large constant in minimization algorith
...
By the way, also use C++17 initialization within “if” statement.
2023-07-04 11:20:51 +02:00
d6ea315a18
Bytecode: further improve some informative messages
2023-06-15 15:37:45 +02:00
192f136b96
Bytecode: improve some informative messages
2023-06-13 16:59:08 +02:00
277c5d4c4f
Bytecode: display real iteration time instead of CPU time
2023-06-13 16:59:08 +02:00
7c9e261355
Bytecode: fix treatment of maxit option
...
It would do one extra iteration.
2023-06-13 16:59:08 +02:00
dcf56b89cd
Bytecode: rework message verbosity
...
Now uses options_.verbosity to decide what to print:
– if options_.verbosity == 0, prints nothing
– if options_.verbosity >= 1, prints iteration counter and duration, and fatal errors
– if options_.verbosity >= 2, additionally print floating point exceptions and
details about algorithmic decisions
2023-06-13 16:59:07 +02:00
73b1850cb5
Bytecode: remove unused function
2023-06-13 16:58:06 +02:00
08d378c244
Bytecode: in evaluate+dynamic mode, return residuals with as many columns as periods
...
Previously it would also include initial and terminal conditions (i.e.
residuals would have periods+maximum_lag+maximum_lead columns). But we do not
care about residuals at the initial and terminal conditions.
This change is for consistency with the perfect_foresight_problem MEX.
2023-06-12 19:18:44 +02:00
90d790f1c3
Bytecode: fix error message
2023-06-12 19:18:44 +02:00
d814a4090b
Bytecode: improve display of residuals error during iterations
2023-06-12 15:12:00 +02:00
4210618dba
🐛 perfect_foresight_problem MEX: check over size of M_.dynamic_g1_sparse_colptr size was incorrect
2023-06-06 16:57:48 +02:00
e86aeaf787
C++20 modernization: use new standard mathematical constants
2023-05-25 19:07:23 +02:00
0f7ab97e69
🐛 local_state_space_iteration_3: real64 kind suffixes were missing for floating-point constants
...
Without the suffix, those constants were interpreted as
real32 (single-precision), hence leading to a loss of precision.
2023-05-24 21:18:23 +02:00
1b02026d1c
Fix GCC 13 warning about overloading that hides virtual methods of base class
2023-05-09 21:36:43 +02:00
aeff2bf48d
Fix GCC 13 warning about ambiguity of operator==() in C++20
...
In C++20, a==b can also be implicitly rewritten as b==a, assuming
commutativity. But if the operator==(const &) is not declared as const, and is
thus asymmetric, this results in an ambiguity because neither of a==b or b==a
is a better match according to the overload resolution rules.
2023-05-09 21:36:41 +02:00
030316f2d8
Bytecode: misc simplifications
2023-04-18 22:25:58 +02:00
afe147d88d
Bytecode: the class dynSparseMatrix is no longer derived from Evaluate
2023-04-18 18:10:45 +02:00
e22972849b
Bytecode: make Evaluate::{symbol_table, steady_state} private and const
2023-04-18 17:46:48 +02:00
4313640f80
Bytecode: allow better compiler optimization of the main evaluation function
...
Use the __restrict__ GCC extension on double* pointers, to avoid aliasing.
NB: the C “restrict” keyword does not exist in C++.
2023-04-18 17:45:15 +02:00
d71db85c71
Bytecode: move more cached information about current block out of the Evaluate class
2023-04-18 17:44:55 +02:00
36e8062e25
Bytecode: remove Evaluate::{EQN_block,EQN_block_number}
2023-04-18 17:37:35 +02:00
d789e6a4c5
Bytecode: move storage for variable and parameter values out of the Evaluate class
2023-04-18 17:37:33 +02:00
d0864689d2
Bytecode: minor simplification related to Evaluate::steady_state
2023-04-18 14:23:26 +02:00
73e4ced39a
Remove more unused Dynare++ code
...
In particular, the libkorder library no longer depends on MatIO.
2023-04-17 18:10:36 +02:00
99cd06c9fd
Move relevant Dynare++ files to mex/sources/libkorder/. Delete the others.
...
Also move some useful documentation to doc/.
2023-04-17 18:10:35 +02:00
36db88c56c
qmc_sequence MEX: use the Fortran version of John Burkardt’s code (instead of C++)
...
Incidentally, this greatly diminishes the compilation time of the MEX.
2023-04-15 16:19:35 +02:00
85351d751c
🐛 qmc_sequence MEX: large input seeds would be truncated
...
The int64 input seed was converted to a double before being converted back to
an int64. But large integers cannot be represented exactly in a double.
2023-04-15 15:48:55 +02:00
9dff1ff28e
🐛 Workaround for buggy int64_T and uint64_T types under Windows with MATLAB < R2015b
...
This bug would impact the qmc_sequence MEX, which manipulates an int64
input/output argument (the seed). Thanks to little-endianness, the bug would
however only manifest for seeds > 2³¹.
2023-04-15 15:48:54 +02:00
e9ecce8293
ricatti_update MEX: remove unused variable
2023-04-15 13:03:43 +02:00
3bf40d92f8
blas_lapack.F08: minor simplification
2023-04-15 12:50:44 +02:00
640be269af
Add “implicit none” statements in BLAS/LAPACK interfaces for safety
2023-04-14 15:08:53 +02:00
62c2881bc2
Fortran 2018 modernization: use the extended “implicit none” syntax
...
The statement “implicit none” forbids implicit typing of variables, but not
implicit procedure declaration. The new “implicit none (type, external)” syntax
forbids both, and is thus safer.
2023-04-14 15:06:37 +02:00
90a408e3c3
Build system: enable the dangling else warning (-Wdangling-else)
...
Adapt the code accordingly.
2023-04-11 17:31:18 +02:00
9f29e2f292
Bytecode: add missing floating point exception handling for sqrt and several trigonometric functions
2023-04-07 17:32:36 +02:00
2740bdee58
Bytecode: remove default clause in switch/case clauses for opcodes
...
This ensures that we will get a compiler warning reminding us to support
opcodes that may be added in the future.
2023-04-07 16:46:35 +02:00
1622c74a53
Bytecode: add missing support for cbrt(), abs() and sign() operators
2023-04-07 16:46:35 +02:00
f27c3abd2f
Bytecode: refactor floating point exception handling
2023-04-07 16:46:32 +02:00
fbaa051489
Bytecode: minor simplification
2023-04-07 10:26:57 +02:00
Sébastien Villemot
Normann Rion