#72: 2nd and 3rd order approximation of purely backward and forward models
--------------------------+-------------------------------------------------
Reporter: michel | Owner:
Type: feature | Status: new
Priority: minor | Milestone: 4.2
Component: Core M-files | Version:
Keywords: |
--------------------------+-------------------------------------------------
- need to write specialized code for 2nd and 3rd order approximation of
purely backward and forward models
--
Ticket URL: <https://www.dynare.org/trac/ticket/72>
Dynare <http://www.dynare.org>
The Dynare project
#117: IRF: document the way they are computed at order 2 and 3
---------------------------+------------------------------------------------
Reporter: sebastien | Owner:
Type: feature | Status: new
Priority: major | Milestone: 4.2
Component: Documentation | Version:
Keywords: |
---------------------------+------------------------------------------------
--
Ticket URL: <https://www.dynare.org/trac/ticket/117>
Dynare <http://www.dynare.org>
The Dynare project
#63: new option instrument
---------------------------+------------------------------------------------
Reporter: michel | Owner: michel
Type: enhancement | Status: new
Priority: major | Milestone:
Component: Documentation | Version:
Keywords: |
---------------------------+------------------------------------------------
Document the new option "instrument" for ramsey_policy in reference manual
and write wiki page on using *_steadystate.m file with ramsey_policy
--
Ticket URL: <https://www.dynare.org/trac/ticket/63>
Dynare <http://www.dynare.org>
The Dynare project
#94: Remove obsolete code dealing with more than one lead and/or one lag
--------------------------+-------------------------------------------------
Reporter: sebastien | Owner:
Type: enhancement | Status: new
Priority: minor | Milestone: 4.2
Component: Core M-files | Version:
Keywords: |
--------------------------+-------------------------------------------------
In the resolution of stochastic models, this code is no longer necessary
since the preprocessor removes any lead and lag of more than one.
But this code should be kept in the deterministic case, since the
preprocessor doesn't do the transformation.
--
Ticket URL: <https://www.dynare.org/trac/ticket/94>
Dynare <http://www.dynare.org>
The Dynare project
#122: Lower and upper bounds on priors used for two purposes
--------------------------+-------------------------------------------------
Reporter: sebastien | Owner:
Type: bug | Status: new
Priority: minor | Milestone: 4.2
Component: Core M-files | Version: 4.1.2
Keywords: |
--------------------------+-------------------------------------------------
The p3 and p4 parameters of estimated_params are currently used for two
distinct purposes:
- specifying the domain of definition of the prior (uniform, generalized
beta and gamma)
- restricting the optimization for the posterior mode on a sub-domain of
the definition of the prior
This double usage of the same values is confusing, and makes some cases
impossible to describe in a MOD file (example: restricting the
optimization on a subdomain of the interval of a uniform).
The fix is probably to distinguish these two usages in the MOD file by
introducing a new syntax for restricting the optimization, and reflecting
that change in the M code.
--
Ticket URL: <https://www.dynare.org/trac/ticket/122>
Dynare <http://www.dynare.org>
The Dynare project
#110: Implement block decomposition for stochastic models
--------------------------+-------------------------------------------------
Reporter: sebastien | Owner: ferhat
Type: enhancement | Status: new
Priority: major | Milestone: 4.2
Component: Core M-files | Version:
Keywords: |
--------------------------+-------------------------------------------------
This can be very useful for some models.
For example, the BoE has a model with bond prices with a horizon of 10
years ahead (i.e. 40 periods). The bond prices do not feedback into the
core of the macro model, so the pricing equations are purely forward.
At order 3, Dynare is currently at pains to solve the model. The block
decomposition would dramatically improve performance here.
--
Ticket URL: <https://www.dynare.org/trac/ticket/110>
Dynare <http://www.dynare.org>
The Dynare project
#121: Revert example MOD files in user guide to mode_compute=4
---------------------------+------------------------------------------------
Reporter: sebastien | Owner:
Type: bug | Status: new
Priority: major | Milestone: 4.2
Component: Documentation | Version:
Keywords: |
---------------------------+------------------------------------------------
The estimation examples in the user guide are not behaving nice with
mode_compute=6, because of the non-deterministic nature of estimation.
Many users were confused because on some runs, the example were failing
because the hessian at the mode was not definite positive.
As a quick fix, the mode_compute option was changed to 6 in
4ec5016e5fed4d51e4ac987c1e8f4bcc85f3148f and
7b59c012b992387c2cc11ab5dd5a8175051cce79.
The problem is that the examples are now a lot slower than before.
We should revert back to mode_compute=4 when #85 is fixed and we have a
way to fix the seed.
--
Ticket URL: <https://www.dynare.org/trac/ticket/121>
Dynare <http://www.dynare.org>
The Dynare project
#127: Output 2nd derivatives of static model when doing identification
--------------------------+-------------------------------------------------
Reporter: sebastien | Owner: sebastien
Type: feature | Status: new
Priority: major | Milestone: 4.3
Component: Preprocessor | Version:
Keywords: |
--------------------------+-------------------------------------------------
This is necessary for global identification.
The code for 2nd static derivatives already exists (used for
planner_objective).
--
Ticket URL: <https://www.dynare.org/trac/ticket/127>
Dynare <http://www.dynare.org>
The Dynare project