December 2010 - Dev - dynare.org

[Trac] #78: Problem with derivatives of power function at point zero
by Trac Dynare 17 Dec '10

17 Dec '10

#78: Problem with derivatives of power function at point zero --------------------------+------------------------------------------------- Reporter: sebastien | Owner: sebastien Type: bug | Status: new Priority: major | Milestone: Component: Preprocessor | Version: 4.1.0 Keywords: | --------------------------+------------------------------------------------- There is currently a problem with the derivatives of the power function f(x)=x^p^ at the point x=0. The power function f(x)=x^p^ well defined for x>0 and for any p, since f(x)=e^p*ln(x)^. The domain of definition of this function can be extended to x=0 with the following rules when p>0: * f(0)=0 * f is differentiable k times at x=0, where k is the integer part of p, and f'^k^=0 And when p is an strictly positive integer, the function is differentiable at any order, and its derivative at zero is zero. When p is a constant, the preprocessor computes the right value of the derivatives at x=0, using simplifications rules. But when p is a parameter, the preprocessor will not give the right value for integer values of p. Currently the derivation rule is applied by the preprocessor, the k-th derivative of f(x) is equal to: p*(p-1)*...*(p-k+1)*x^p-k^ So when p is an integer, the (p+1)-th derivative computed at x=0 will give a NaN (0 times infinity), while it ought to be zero. This problem typically arises in models with adjustment costs where the curvature (the exponent) of the cost is a integer parameter, and the cost is zero at steady state. The solution is probably to create, in the preprocessor, a new operator called "derivative of power function at order k". This operator would output a code like that in the M-file: {{{ dxp=deriv(x,p,k) % The k-th derivative of x^p if (p > 0) && (abs(p - round(p)) < 1e-12 && k >= p dxp = 0; else dxp = x^(p-k); for i=1:k-1 dxp = dxp*p; p = p-1; end end }}} -- Ticket URL: <https://www.dynare.org/trac/ticket/78> Dynare <http://www.dynare.org> The Dynare project

1 7

[Trac] #156: distclean doesn't work if Dynare was built with --disable-octave
by Trac Dynare 15 Dec '10

15 Dec '10

#156: distclean doesn't work if Dynare was built with --disable-octave --------------------------+------------------------------------------------- Reporter: michel | Owner: sebastien Type: bug | Status: new Priority: minor | Milestone: Component: Preprocessor | Version: Keywords: | --------------------------+------------------------------------------------- This seems to be a problem with the options of the build system. -- Ticket URL: <https://www.dynare.org/trac/ticket/156> Dynare <http://www.dynare.org> The Dynare project

1 1

Re: [DynareDev] Fwd: Re: SLICOT Library
by stephane.adjemian＠gmail.com 14 Dec '10

14 Dec '10

Hi George, I plan to (for dynare>=4.3) entirely rewrite the kalman filter routines (matlab codes) to allow for unexpected structural breaks. I use Slicot because this library provides a lot of utilities for handling state space models and evaluating kalman filters. For instance, I use a solver for the discrete Riccatti equations, and use the solution for initializing the kalman filter. This gives a much more stable evaluation of the likelihood (also the evaluation is much faster). This evaluation is biased but its variance is smaller. As a first approximation (for large models) I use this trick when looking for the posterior mode (or maximum likelihood). This gives a very good initial guess when estimating the posterior mode. I will also use the library to reduce the size of the state space model and for comparison purposes. Best, Stéphane. George Perendia <george(a)perendia.orangehome.co.uk> writes: > Hi Stephane > From the email below I gather you are planning to use SLICOT for some work with KF. > I was just intrigued and wandered if you can shed me some light on that or send me some existing paper-work if poss and you have a moment for that. > > Best regards > George > > -------- Original Message -------- > Subject: Re: SLICOT Library > Date: Mon, 13 Dec 2010 10:24:07 +0100 > From: Sébastien Villemot <sebastien.villemot(a)ens.fr> > To: George Perendia <george(a)perendia.orangehome.co.uk> > > George Perendia <george(a)perendia.orangehome.co.uk> writes: > >> I noticed a new sub-directory of Dynare mex files - SLICOT Library. >> What is going to be the main use of the control system SLICOT Library within >> Dynare that is in mex slicot directory and how will it be accessed? Are there >> any wiki pages on it? > > SLICOT is a multi-purpose optimal control library written in Fortran, > see http://www.slicot.org/ > > Stéphane is now using it in some extensions of the Kalman filter. If you > want more details on that, please ask him directly. > > We included it in Dynare distribution (instead of simply making Dynare > depend on it), because there is no pre-packaged version of this library > (neither in Debian/Ubuntu nor in Cygwin). > > Best, -- Stéphane Adjemian Université du Maine, GAINS & CEPREMAP http://www.dynare.org/stepan

1 0

[Trac] #154: Change build process to make libslicot optional on Mac OS X
by Trac Dynare 13 Dec '10

13 Dec '10

#154: Change build process to make libslicot optional on Mac OS X --------------------------------------+------------------------------------- Reporter: houtanb | Owner: sebastien Type: bug | Status: new Priority: minor | Milestone: 4.2 Component: Build system / Packaging | Version: Keywords: | --------------------------------------+------------------------------------- XCode doesn't provide a Fortran compiler so Mac users will either have to find one precompiled or compile it on their own. Hence, libslicot should be optional in the build process. -- Ticket URL: <https://www.dynare.org/trac/ticket/154> Dynare <http://www.dynare.org> The Dynare project

1 2

[Trac] #155: GraphvizDigraph has been removed from recent Boost versions
by Trac Dynare 13 Dec '10

13 Dec '10

#155: GraphvizDigraph has been removed from recent Boost versions --------------------------+------------------------------------------------- Reporter: michel | Owner: ferhat Type: bug | Status: new Priority: critical | Milestone: Component: Preprocessor | Version: Keywords: | --------------------------+------------------------------------------------- It appears that GraphvizDigraph was removed from Boost in version 1.43. See https://svn.boost.org/trac/boost/ticket/4762 GrahphvizDigraph is used in MinimumFeedbackSet.hh (line 58) and in ModelTree.cc (line 513) Boost version 1.45 is available in Redhat version 6, but only version 1.42 is available in Debian testing/unstable -- Ticket URL: <https://www.dynare.org/trac/ticket/155> Dynare <http://www.dynare.org> The Dynare project

1 2

Re: [DynareDev] Problems building Dynare on RH ES4
by Michel Juillard 10 Dec '10

10 Dec '10

Dear Paul, thank you very much for reporting the problem. It appears that GraphvizDigraph was removed from Boost in version 1.43. See https://svn.boost.org/trac/boost/ticket/4762 As Debian is still at version 1.42, we weren't aware from the problem. We are going to fix this. In the meantime, you should try to get version 1.42 of the Boost graphic library. Best, Michel PS. The Dynare forum is a better way to contact us, rather than the email of the webmaster. m On 12/10/2010 04:49 PM, michaud(a)umich.edu wrote: > You are receiving this mail because Paul Michaud > michaud(a)umich.edu > is sending feedback about the site administered by you at http://www.dynare.org. > The message sent was: > > Hi, I'm attempting to build Dynare 4.1.3 from source in a Red Hat Enterprise Linux ES release 4 (Nahant Update 8) environment. To configure I use the following command: ./configure --prefix=/software/dynare/v4.1.3 --with-boost=/software/Boost/v1.45.0 --with-matlab=/software/matlab_2009b MATLAB_VERSION=R2009b According to the configure logs a "make" should build the Dynare preprocessor. Said "make" builds the libmicro.a library just fine. However, in the preprocessor build step it chokes building dynare_m-ModelTree.o. It's complaint is that (starting on line 58) in the MinimumFeedbackSet.hh include file (used by ModelTree.cc) GraphvizDigraph is not declared. The same issue comes up on line 513 of ModelTree.cc. There are some other bits as well but this appears to be the main catalyst. I can forward you the make output but didn't want to clutter up this initial email with all that goo. :-) Thanks much for any help that you can provide! :-) Paul ============================================================================ Paul Michaud Research Computing Stephen M. Ross School of Business at the University of Michigan Email: michaud(a)umich.edu Phone: 734-763-5677 Fax: 734-936-6029 ============================================================================ > > >

1 0

Re: [DynareDev] [Trac] #9: Add derivatives of external functions in the model
by Trac Dynare 10 Dec '10

10 Dec '10

#9: Add derivatives of external functions in the model ---------------------------+------------------------------------------------ Reporter: sebastien | Owner: ferhat Type: feature | Status: closed Priority: minor | Milestone: 4.2 Component: Preprocessor | Version: Resolution: fixed | Keywords: ---------------------------+------------------------------------------------ Changes (by sebastien): * status: assigned => closed * resolution: => fixed Comment: Implemented in bytecode in 97664607b2f839caa2e733d72655307dc6023ea0 -- Ticket URL: <https://www.dynare.org/trac/ticket/9#comment:12> Dynare <http://www.dynare.org> The Dynare project

1 0

[Trac] #73: The macroprocessor fails when the file ends with @#endif or @#endfor without a new line at the EOF
by Trac Dynare 06 Dec '10

06 Dec '10

#73: The macroprocessor fails when the file ends with @#endif or @#endfor without a new line at the EOF --------------------------+------------------------------------------------- Reporter: sebastien | Owner: sebastien Type: bug | Status: new Priority: minor | Milestone: Component: Preprocessor | Version: 4.1.0 Keywords: | --------------------------+------------------------------------------------- -- Ticket URL: <https://www.dynare.org/trac/ticket/73> Dynare <http://www.dynare.org> The Dynare project

1 2

Re: [DynareDev] [Trac] #22: Expand specification of priors (i.e. priors on relative variances between 2 shocks)
by Trac Dynare 03 Dec '10

03 Dec '10

#22: Expand specification of priors (i.e. priors on relative variances between 2 shocks) ---------------------------+------------------------------------------------ Reporter: sebastien | Owner: michel Type: feature | Status: assigned Priority: major | Milestone: 4.3 Component: Core M-files | Version: Resolution: | Keywords: ---------------------------+------------------------------------------------ Changes (by sebastien): * owner: => michel * status: new => assigned * milestone: 4.2 => 4.3 -- Ticket URL: <https://www.dynare.org/trac/ticket/22#comment:3> Dynare <http://www.dynare.org> The Dynare project

1 0

Re: [DynareDev] [Trac] #17: Update user guide
by Trac Dynare 03 Dec '10

03 Dec '10

#17: Update user guide ----------------------------+----------------------------------------------- Reporter: sebastien | Owner: michel Type: enhancement | Status: assigned Priority: major | Milestone: 4.3 Component: Documentation | Version: Resolution: | Keywords: ----------------------------+----------------------------------------------- Changes (by sebastien): * owner: tommaso => michel * status: new => assigned * milestone: 4.2 => 4.3 -- Ticket URL: <https://www.dynare.org/trac/ticket/17#comment:4> Dynare <http://www.dynare.org> The Dynare project

1 0