Marco Ratto <marco.ratto@jrc.ec.europa.eu> writes:

thanks a lot for the fix. I think this is OK. Currently, to get the derivative of static model w.r.t. parameters, I use the derivatives of the dynamic model using the steady state for all leads and lags. So, in principle, it is already available and, in combination with the Jacobian, they provide the derivatives of steady state w.r.t. params.

Of course, it would be also helpful to write an explicit function for derivatives of the steady state model, if this is not too much of a burden for you. However, this does not seem a 'blocking' issue for my development, unless I have skipped something relevant in this sense.

Ok, then I leave it as it is (for release 4.2), and I open a ticket for the derivatives of the static model (for release 4.3). Best,

On 1/13/2011 6:12 PM, Trac Dynare wrote:

...

#128: Fix derivatives of STEADY_STATE operator w.r.t. parameters in dynamic model ---------------------------+------------------------------------------------ Reporter: sebastien | Owner: sebastien Type: bug | Status: new Priority: major | Milestone: 4.2 Component: Preprocessor | Version: Resolution: | Keywords: ---------------------------+------------------------------------------------

Comment(by sebastien):

Commit 9c6d65bc0a4742f0cc30eaa129567d11a5fc12a3 fixes this.

Two new input arguments are added to the params_deriv file: - ss_param_deriv: matrix for derivatives of steady state w.r.t. parameters (rows: endogenous, columns: parameters) - ss_param_2nd_deriv: 3D array for 2nd derivatives of steady state w.r.t. parameters (1D: endogenous, 2D and 3D: parameters)

Marco: can you confirm that this is enough for fixing the problem? I suspect that you also need the derivatives of the static model w.r.t. parameters.

_______________________________________________ Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev

-- Sébastien Villemot CEPREMAP - Paris School of Economics Homepage: http://www.dynare.org/sebastien Phone: +33 1 40 77 49 90 PGP Key: 0xA6C029B9D06B2913D71C105EBE37E801FB6EFF8B (http://pgp.mit.edu/)