July 2012 - Dev - dynare.org

[Trac] #209: check use of set_default_option()
by Trac Dynare 06 Dec '12

06 Dec '12

#209: check use of set_default_option() --------------------------+------------------------------------------------- Reporter: michel | Owner: Type: bug | Status: new Priority: major | Milestone: 4.3 Component: Core M-files | Version: Keywords: | --------------------------+------------------------------------------------- set_default_option takes an action only if the field name doesn't exist in options_. But, now, most options have a default value set in global_initialization.m, so most of the calls to set_default_option are useless and may, sometimes, not produce the desired results. -- Ticket URL: <https://www.dynare.org/trac/ticket/209> Dynare <http://www.dynare.org> The Dynare project

1 4

[Trac] #262: unify names for maxit options
by Trac Dynare 05 Dec '12

05 Dec '12

#262: unify names for maxit options ---------------------------------+------------------------------------------ Reporter: michel | Owner: michel Type: enhancement | Status: new Priority: major | Milestone: 5.0 Component: Block decomposition | Version: Keywords: | ---------------------------------+------------------------------------------ Several functions need an optional maximum number of iterations. There is currently no unified name style for these options. Until now, we tried to have exactly the same name for the option in the language and the options_ fieldname. In that case, I wonder if we should not have always maxit in the language and specialized the fieldname with a reference to the function. For example maxit in steady -> options_.solve_maxit -- Ticket URL: <https://www.dynare.org/trac/ticket/262> Dynare <http://www.dynare.org> The Dynare project

1 1

[Trac] #160: Add derivatives of static model w.r.t. parameters
by Trac Dynare 29 Nov '12

29 Nov '12

#160: Add derivatives of static model w.r.t. parameters --------------------------+------------------------------------------------- Reporter: sebastien | Owner: sebastien Type: enhancement | Status: new Priority: minor | Milestone: 4.3 Component: Preprocessor | Version: Keywords: | --------------------------+------------------------------------------------- Currently, the derivatives of the static model w.r.t. parameters are computed using the derivatives of the dynamic model w.r.t. parameters. It would be more convenient to provide directly these derivatives. -- Ticket URL: <https://www.dynare.org/trac/ticket/160> Dynare <http://www.dynare.org> The Dynare project

1 2

[Trac] #263: latex variable with exponent causes compilation error
by Trac Dynare 29 Nov '12

29 Nov '12

#263: latex variable with exponent causes compilation error --------------------------+------------------------------------------------- Reporter: houtanb | Owner: sebastien Type: bug | Status: new Priority: minor | Milestone: 5.0 Component: Preprocessor | Version: 4.3.0 Keywords: | --------------------------+------------------------------------------------- See: http://www.dynare.org/phpBB3/viewtopic.php?f=1&t=3846 Quick fix is to require users to place brackets around latex variables containing exponents (and probably underscores) Complete fix is to place braces around variables with lead/lags and those variables with no lead/lag but that are followed by an exponent. In the user's case, we would output: {B^{problem}_{t-1}}^{\rho} -- Ticket URL: <https://www.dynare.org/trac/ticket/263> Dynare <http://www.dynare.org> The Dynare project

1 2

[Trac] #267: design an interface for assigning moment/irf contraints (e.g. sign constraints)
by Trac Dynare 29 Nov '12

29 Nov '12

#267: design an interface for assigning moment/irf contraints (e.g. sign constraints) -----------------------------------------------------+---------------------- Reporter: ratto | Owner: sebastien Type: feature | Status: new Priority: major | Milestone: 5.0 Component: Identification and sensitivity analysis | Version: Keywords: | -----------------------------------------------------+---------------------- Currently there is no way of setting a set of constraints to irfs or moments of DSGE model to feed to some calibration procedure like Monte Carlo filtering in the sensitivity analysis toolbox. Indeed the threshold_redform option in sensitivity analysis would only allow to set, e.g., the SAME sign constraint to a set of irfs, which is not very handy. So, we may think to the following syntax for sign constraints in moments: moment_calibration; y, y(-1), +; [i.e. auto-correlation with 1 lag] c, g(-3), +; [i.e. auto-correlation with generic lags] i, g, -; [i.e. cross-correlation, contemporaneous] end; for irfs [y c i are endogenous, em eg are exogenous shocks]: irf_calibration; y, em, -; c, eg, +; i, eg, -; end; or the following for more detailed numeric boundaries, in principle mixed with simple sign constraints moment_calibration; y, y(-1), [0.5 1]; [auto-correlation with 1 lag] c, g(-3), +; [auto-correlation with generic lags] i, g, [-1, -0.5]; [cross-correlation, contemporaneous] end; irf_calibration; y, em, [-0.01 0]; c, eg, [0 0.02]; i, eg, -; end; The sign constraint can also be defined using [0 inf] for + [-inf 0] for - -- Ticket URL: <https://www.dynare.org/trac/ticket/267> Dynare <http://www.dynare.org> The Dynare project

1 2

[Trac] #205: add interface and documentation to model_diagnostics
by Trac Dynare 16 Nov '12

16 Nov '12

#205: add interface and documentation to model_diagnostics --------------------------+------------------------------------------------- Reporter: sebastien | Owner: sebastien Type: enhancement | Status: new Priority: major | Milestone: 4.3 Component: Preprocessor | Version: Keywords: | --------------------------+------------------------------------------------- -- Ticket URL: <https://www.dynare.org/trac/ticket/205> Dynare <http://www.dynare.org> The Dynare project

1 3

[Trac] #141: Replace strmatch which is a deprecated function in Matlab
by Trac Dynare 16 Nov '12

16 Nov '12

#141: Replace strmatch which is a deprecated function in Matlab --------------------------+------------------------------------------------- Reporter: houtanb | Owner: Type: bug | Status: new Priority: major | Milestone: Component: Core M-files | Version: Keywords: | --------------------------+------------------------------------------------- Use strncmp instead, which has a different return type http://www.mathworks.com/help/techdoc/ref/strmatch.html -- Ticket URL: <https://www.dynare.org/trac/ticket/141> Dynare <http://www.dynare.org> The Dynare project

1 7

[Trac] #266: allow inf in threshold redform option of sensitivity analysis
by Trac Dynare 16 Nov '12

16 Nov '12

#266: allow inf in threshold redform option of sensitivity analysis -----------------------------------------------------+---------------------- Reporter: ratto | Owner: ratto Type: enhancement | Status: new Priority: major | Milestone: 5.0 Component: Identification and sensitivity analysis | Version: 4.3.0 Keywords: | -----------------------------------------------------+---------------------- Currently, only real numbers are accepted by the option threshold_redform of the sensitivity toolbox. However, it would be handful to allow inf as input, e.g. if one wants to filter positive numbers in an irf, one would like to use a syntax [0 inf]. -- Ticket URL: <https://www.dynare.org/trac/ticket/266> Dynare <http://www.dynare.org> The Dynare project

1 2

[Trac] #261: Option to require that all endo/exo are initialized in initval
by Trac Dynare 07 Nov '12

07 Nov '12

#261: Option to require that all endo/exo are initialized in initval --------------------------+------------------------------------------------- Reporter: sebastien | Owner: sebastien Type: feature | Status: new Priority: major | Milestone: 5.0 Component: Preprocessor | Version: Keywords: | --------------------------+------------------------------------------------- Implement a new option to initval command, "all_values_required". The preprocessor would stop and reports which endogenous or exogenous variable has not been explicitly initialized if any (instead of using a default zero value). Request from the IMF for GIMF -- Ticket URL: <https://www.dynare.org/trac/ticket/261> Dynare <http://www.dynare.org> The Dynare project

1 4

[Trac] #240: rewrite getPowerDeriv assignments in temporary terms
by Trac Dynare 30 Oct '12

30 Oct '12

#240: rewrite getPowerDeriv assignments in temporary terms --------------------------+------------------------------------------------- Reporter: houtanb | Owner: houtanb Type: bug | Status: new Priority: major | Milestone: 4.3 Component: Preprocessor | Version: Keywords: | --------------------------+------------------------------------------------- From Stephane: In models with power functions of the form x**a (utility functions, production functions, agregation functions, quadratic costs, ...) the evaluation of the jacobian matrix (or higher order derivates) necessitate potentially a huge number of calls to the routine getPowerDeriv (the name of the routine is explicit enough). In my example (a real business cycle model with perfect foresight) most of the time is spent in this routine, while, if my understanding is correct, this is not necessary because x>0. I wrote a mex file (see the new branch called mex-!GetPowerDeriv) as a replacement for the matlab routine, unfortunatly the overhead cost is so high that the use of the mex file increases the total time of execution! In my case I resolved the issue by using option use_dll, but this is not a solution for the majority of our users (they would have to install cygwin/gcc and configure matlab/mex). So my question is: Do we really need to call this routine in all situations ? Would it not be possible to replace calls to this routine by something like: if (abs(x)>1e-12) tmp = ANALYTICAL_EXPRESSION_OF_THE_DERIVATIVE{x,params}; else tmp = 0; end -- Ticket URL: <https://www.dynare.org/trac/ticket/240> Dynare <http://www.dynare.org> The Dynare project

1 3