When computing the derivatives or block decomposition of the planner objective,
the epilogue or the original Ramsey model, the preprocessor would talk about
dynamic/static model, which was confusing. It now uses the right terminology.
If block decomposition fails, error out if “block” option was passed, but not
otherwise.
This commit does not modify the generated files.
This is a preliminary step for dynare#1859.
– factorize common code between the static and the dynamic version
– reorganise language-specific code into dedicated functions
– use a function template in the main helper to do some computations
at compile-time (using constexpr features)
– a generic one: CommonEnums.hh
– and a bytecode-specific one: Bytecode.hh
By the way, rename global constant “near_zero” into “power_deriv_near_zero”,
for clarity.
In particular, use this feature in many loops which feature a special treatment
for the first iteration, using a boolean variable (replacing iterator
manipulation). By the way, also use std::exchange() to simultaneously test the
value of this variable and update it.
– Indicate whether we are trying to normalize the static or dynamic model
– If failed to normalize the static model, suggest to use the “no_static”
option
– Remove a superfluous error message
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.
Consequently drop “occbin” option to “model”.
Incidentally, allow more values in equation tag names (previously some keywords
such as “alpha” were disallowed).
Ref. #68
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
— 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
— 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
Some evaluated blocks could be incorrectly merged, leading to wrong
results (e.g. blocks incorrectly marked as evaluable forward).
Bug introduced in 39407083be
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.
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.
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.
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.
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).
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).
This is for consistency with other graph algorithms. Previously it would return
an index from vertex_index1.
See also 1d92adacf4 (which this commit basically
reverts).
Also do some cosmetic changes in calling graph code.
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).
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.
— 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.
Since bef537d40a, constant equations were not
simplified as soon as they had a tag attached.
But this is too wide a restriction. In particular, this breaks the trend
component models which have a target that is set to a constant.
So we now only skip the replacement in the case where there is an “mcp” tag.
Ref. dynare#1697
This was only adding unneeded complexity, for no clear reason (we’re very far
from reaching 2³¹ equations, and if we wanted to support models that large, it
would be better to use long integers to avoid being limited to 2³²).
Sébastien Villemot
MichelJuillard
Willi Mutschler