Dev May 2009

dev@dynare.org

6 participants
14 discussions

model derivatives with respect to parameters
by Michel Juillard 19 Aug '09

19 Aug '09

Dear Nikolay, we have started to work on producing the derivatives with respect to parameters. We are planing to create a new function, that will return a 3-dimensional array, by equation, variable and parameter. You could then select one parameter and squeeze the array in the parameter dimension to obtain a matrix of Jacobian derivatives with respect to a given parameter. Do you see any problem with this implementation? We should be able to provide in about two weeks. All the best, Michel

4 6

Re: [DynareDev] Kalman Filter
by Michel Juillard 29 May '09

29 May '09

On closer inspection, I don't think that the expression pointed by George in kalman_filter.m is wrong: 1. reste = smpl-t or the number of periods during which the filter is stationary. This shouldn't be larger than T-start+1 2. it is problematic (see below) but not wrong to add all the determinants at once in the last period of the stationary filter 3. I don't think this explains the difference with the C++ version of the filter and we still have to look for it. 4. it remains that the current code is very unclear and that if LIK is correct the vector lik doesn't have the correct constants on each elements. 5. I would like to simplify the code and add the correct constant to each element of the lik vector. It would be a little bit less efficient in Matlab than the current code, but I doubt it would be noticeable. Stephane, what do you think? Best Michel G. Perendia wrote: > Dear Michel > > I think I found an error in Dynare Matlab kalman_filter. suite of utilities > which affects the likelihood LIK results with start>1 (i.e. presampling>0): > > the calculation speed-up construct which relies on converged covariance > matrix > > lik(t) = lik(t) + reste*log(dF); > > adds reste * log(dF) to the last-1 (i.e. the smpl) member of lik > (the last, the lik(smpl+1) one contains smpl*pp*log(2*pi)) > but reste is usually larger than T-start+1 so that > > LIK = .5*(sum(lik(start:end))-(start-1)*lik(smpl+1)/smpl) > > has much more log(dF)s added than required since they are all concentrated > in the last-1 (the T) member > > For example, if I change the above construct to > lik(t) = lik(t) + min(reste,(smpl-start+1))*log(dF); > > the reported likelihood for presample=40 from Matlab KF is > 1640935.5855267849 > which is nearly the same as that from C++ KF below: > 1640935.5854489324 > > Shall I make changes to kalman/likelihood/ KFs and upload the .m files? > This problem affects also the older versions of DiffuseLikelihood**.m too. > > Best regards > > George > artilogica(a)btconnect.com > > ----- Original Message ----- > From: "Michel Juillard" <michel.juillard(a)ens.fr> > To: "G. Perendia" <george(a)perendia.orangehome.co.uk> > Sent: Tuesday, May 26, 2009 10:32 AM > Subject: Re: Kalman Filter+PS > > > >> Hi George, >> >> >> >>> Re 1) below: >>> I modified C++ KF so that it reports log-likelihood for given >>> start/preampling in same/similar manner as the Matlab KFs do and I am >>> getting approximately close results, e.g. >>> >>> ll= -1640935.5854489324 for C++ and >>> (-) 1640482.4179242959 for Matlab KF (for start=41, i.e. >>> > presample=40). > >>> whilst they appear same for presample=0 (e.g.2.5906e+006), i.e. >>> -2590556.989730841 vs >>> 2590556.989778722 >>> >>> Are those results acceptably close or should I investigate further where >>> > the > >>> above difference may come form? >>> >>> >>> >> This indicates a problem . The difference should be the same with and >> without presample. It may come from the computation of the likelihood >> constant. This is done in a very obscure manner in Dynare Matlab. >> > > > >

4 18

Re: [DynareDev] Kalman Filter
by G. Perendia 28 May '09

28 May '09

PS: Aim is to emulate calling protocol for current matlab/kalman/likelihood .m files. I am currently adjusting the kalman_filter.cpp dll suite (kalman_filter.cpp to start with) to create internally H and Pinf 0 matrices of suitable size if empty or single 0 passed to them and so that non-diffuse on can take mf instead Z as C++ currently expects. (The matlab\*.cpp files have already been adjusted to return likelihood array as .m files do.) in meantime one may use .m file to prepare those on the lines of the kalmandll_test.m I just enclosed in the matlab subdirectory of sources/kalman repository. Best regards George ----- Original Message ----- From: G. Perendia To: List for Dynare developers Sent: Wednesday, May 27, 2009 2:56 PM Subject: Re: [DynareDev] Kalman Filter The filter routines there have names and should follow calling rules for the par matlab/kalman/likelihood ones but they do not yet take mf and still expect Z matrix of 1s to map system to observables instead and Pinf must not be empty but at least be initialised to a zeros matrix of Pstar size. The smoother is generic as it was, not suitable for running at all Best regards George ----- Original Message ----- From: Stéphane Adjemian To: List for Dynare developers Sent: Wednesday, May 27, 2009 2:46 PM Subject: Re: [DynareDev] Kalman Filter Hi George 2009/5/27 G. Perendia <george(a)perendia.orangehome.co.uk> Hi The C++ Kalman dll driver routines in /mex/kalman/matlab are still in development Sure... and the code there is still subject to developer testing ... But I'd like to play with it. and the Makefile there is not yet complete but it should use blas/lapack from Matlab libraries. Ok. I will adapt your Makefile for my platform. Best, Stéphane. Best regards George Perendia Tel.: 02072815392 Mob: 07951415480 ----- Original Message ----- From: Stéphane Adjemian To: List for Dynare developers Sent: Wednesday, May 27, 2009 2:18 PM Subject: Re: [DynareDev] Kalman Filter Oui, oui... I never use the presampling option... And with all the examples I have ever considered it takes much more than 10 iterations to get to the steady state kalman filter. By the way, is there somewhere linux a Makefile for the cc kalman routines? Also, why do we need Atlas (if we also use matlab's lapack/blas libraries)? Stéphane. 2009/5/27 Michel Juillard <michel.juillard(a)ens.fr> Si le filtre converge a l'etat stationnaire a une date avant start (la date a partir de laquelle on cumule la vraisemblance), alors additionner tous les determinants des iterations dans le filtre stationnaire a la derniere periode fausse le calcul de la constante. amicalement Michel Stéphane Adjemian wrote: Thanks Michel. Your commit is ok for me. I do not yet understand the problem raised by George. I need to go through his example... Stéphane. 2009/5/27 Michel Juillard <michel.juillard(a)ens.fr <mailto:michel.juillard@ens.fr>> Thanks Stephane, I just uploaded a new version of kalman_filter.m Tell me what you think Best Michel Stéphane Adjemian wrote: Hi all, I agree, the matlab code is very unclear (even if I had fun writting it this way ;-) and prone to errors if one uses the vector lik (Marco is using it). I would rather prefer to add the constants outside of the loop with a (sub)vector operation, this should be more efficient. I will do it today or tomorrow. Best, Stéphane. 2009/5/27 Michel Juillard <michel.juillard(a)ens.fr <mailto:michel.juillard@ens.fr> <mailto:michel.juillard@ens.fr <mailto:michel.juillard@ens.fr>>> On closer inspection, I don't think that the expression pointed by George in kalman_filter.m is wrong: 1. reste = smpl-t or the number of periods during which the filter is stationary. This shouldn't be larger than T-start+1 2. it is problematic (see below) but not wrong to add all the determinants at once in the last period of the stationary filter 3. I don't think this explains the difference with the C++ version of the filter and we still have to look for it. 4. it remains that the current code is very unclear and that if LIK is correct the vector lik doesn't have the correct constants on each elements. 5. I would like to simplify the code and add the correct constant to each element of the lik vector. It would be a little bit less efficient in Matlab than the current code, but I doubt it would be noticeable. Stephane, what do you think? Best Michel G. Perendia wrote: Dear Michel I think I found an error in Dynare Matlab kalman_filter. suite of utilities which affects the likelihood LIK results with start>1 (i.e. presampling>0): the calculation speed-up construct which relies on converged covariance matrix lik(t) = lik(t) + reste*log(dF); adds reste * log(dF) to the last-1 (i.e. the smpl) member of lik (the last, the lik(smpl+1) one contains smpl*pp*log(2*pi)) but reste is usually larger than T-start+1 so that LIK = .5*(sum(lik(start:end))-(start-1)*lik(smpl+1)/smpl) has much more log(dF)s added than required since they are all concentrated in the last-1 (the T) member For example, if I change the above construct to lik(t) = lik(t) + min(reste,(smpl-start+1))*log(dF); the reported likelihood for presample=40 from Matlab KF is 1640935.5855267849 which is nearly the same as that from C++ KF below: 1640935.5854489324 Shall I make changes to kalman/likelihood/ KFs and upload the .m files? This problem affects also the older versions of DiffuseLikelihood**.m too. Best regards George artilogica(a)btconnect.com <mailto:artilogica@btconnect.com> <mailto:artilogica@btconnect.com <mailto:artilogica@btconnect.com>> ----- Original Message ----- From: "Michel Juillard" <michel.juillard(a)ens.fr <mailto:michel.juillard@ens.fr> <mailto:michel.juillard@ens.fr <mailto:michel.juillard@ens.fr>>> To: "G. Perendia" <george(a)perendia.orangehome.co.uk <mailto:george@perendia.orangehome.co.uk> <mailto:george@perendia.orangehome.co.uk <mailto:george@perendia.orangehome.co.uk>>> Sent: Tuesday, May 26, 2009 10:32 AM Subject: Re: Kalman Filter+PS Hi George, Re 1) below: I modified C++ KF so that it reports log-likelihood for given start/preampling in same/similar manner as the Matlab KFs do and I am getting approximately close results, e.g. ll= -1640935.5854489324 for C++ and (-) 1640482.4179242959 for Matlab KF (for start=41, i.e. presample=40). whilst they appear same for presample=0 (e.g.2.5906e+006), i.e. -2590556.989730841 vs 2590556.989778722 Are those results acceptably close or should I investigate further where the above difference may come form? This indicates a problem . The difference should be the same with and without presample. It may come from the computation of the likelihood constant. This is done in a very obscure manner in Dynare Matlab. _______________________________________________ Dev mailing list Dev(a)dynare.org <mailto:Dev@dynare.org> <mailto:Dev@dynare.org <mailto:Dev@dynare.org>> http://www.dynare.org/cgi-bin/mailman/listinfo/dev -- Stéphane Adjemian CEPREMAP & Université du Maine Tel: (33)1-43-13-62-39 ------------------------------------------------------------------------ _______________________________________________ Dev mailing list Dev(a)dynare.org <mailto:Dev@dynare.org> http://www.dynare.org/cgi-bin/mailman/listinfo/dev _______________________________________________ Dev mailing list Dev(a)dynare.org <mailto:Dev@dynare.org> http://www.dynare.org/cgi-bin/mailman/listinfo/dev -- Stéphane Adjemian CEPREMAP & Université du Maine Tel: (33)1-43-13-62-39 ------------------------------------------------------------------------ _______________________________________________ Dev mailing list Dev(a)dynare.org http://www.dynare.org/cgi-bin/mailman/listinfo/dev _______________________________________________ Dev mailing list Dev(a)dynare.org http://www.dynare.org/cgi-bin/mailman/listinfo/dev -- Stéphane Adjemian CEPREMAP & Université du Maine Tel: (33)1-43-13-62-39 ------------------------------------------------------------------------ _______________________________________________ Dev mailing list Dev(a)dynare.org http://www.dynare.org/cgi-bin/mailman/listinfo/dev _______________________________________________ Dev mailing list Dev(a)dynare.org http://www.dynare.org/cgi-bin/mailman/listinfo/dev -- Stéphane Adjemian CEPREMAP & Université du Maine Tel: (33)1-43-13-62-39 ---------------------------------------------------------------------------- _______________________________________________ Dev mailing list Dev(a)dynare.org http://www.dynare.org/cgi-bin/mailman/listinfo/dev

2 4

Code beautification/indentation
by Sébastien Villemot 27 May '09

27 May '09

Hi everyone, I have found a great tool for indenting/beautifying C++ code: uncrustify I added a config file for it in the SVN, and some doc on the wiki (page CodingStandards). As an example, I have beautified korderpert code using that tool, and it works great. Best, -- Sébastien Villemot

2 1

Kordpert
by G. Perendia 27 May '09

27 May '09

Michel We should not use Dyn_pp subdirectory in korderpert directory of mex/sources for building dynarelib.a needed for korderpert but up-to-date dynare++. It remained there from the initial plan to separate two developments and may not be up-to-date.. If that is OK, I will delete Dyn_pp from SVN Best regards George artilogica(a)btconnect.com

2 1

Re: [DynareDev] Kalman Filter
by Michel Juillard 27 May '09

27 May '09

Thank you very much George. I will take a look at it. I think that the correction should be more direct than the min() construct Best Michel G. Perendia wrote: > Dear Michel > > I think I found an error in Dynare Matlab kalman_filter. suite of utilities > which affects the likelihood LIK results with start>1 (i.e. presampling>0): > > the calculation speed-up construct which relies on converged covariance > matrix > > lik(t) = lik(t) + reste*log(dF); > > adds reste * log(dF) to the last-1 (i.e. the smpl) member of lik > (the last, the lik(smpl+1) one contains smpl*pp*log(2*pi)) > but reste is usually larger than T-start+1 so that > > LIK = .5*(sum(lik(start:end))-(start-1)*lik(smpl+1)/smpl) > > has much more log(dF)s added than required since they are all concentrated > in the last-1 (the T) member > > For example, if I change the above construct to > lik(t) = lik(t) + min(reste,(smpl-start+1))*log(dF); > > the reported likelihood for presample=40 from Matlab KF is > 1640935.5855267849 > which is nearly the same as that from C++ KF below: > 1640935.5854489324 > > Shall I make changes to kalman/likelihood/ KFs and upload the .m files? > This problem affects also the older versions of DiffuseLikelihood**.m too. > > Best regards > > George > artilogica(a)btconnect.com > > ----- Original Message ----- > From: "Michel Juillard" <michel.juillard(a)ens.fr> > To: "G. Perendia" <george(a)perendia.orangehome.co.uk> > Sent: Tuesday, May 26, 2009 10:32 AM > Subject: Re: Kalman Filter+PS > > > >> Hi George, >> >> >> >>> Re 1) below: >>> I modified C++ KF so that it reports log-likelihood for given >>> start/preampling in same/similar manner as the Matlab KFs do and I am >>> getting approximately close results, e.g. >>> >>> ll= -1640935.5854489324 for C++ and >>> (-) 1640482.4179242959 for Matlab KF (for start=41, i.e. >>> > presample=40). > >>> whilst they appear same for presample=0 (e.g.2.5906e+006), i.e. >>> -2590556.989730841 vs >>> 2590556.989778722 >>> >>> Are those results acceptably close or should I investigate further where >>> > the > >>> above difference may come form? >>> >>> >>> >> This indicates a problem . The difference should be the same with and >> without presample. It may come from the computation of the likelihood >> constant. This is done in a very obscure manner in Dynare Matlab. >> > > > >

1 0

Re: [DynareDev] quasi-triangular matrices
by Michel Juillard 26 May '09

26 May '09

Great! Then take contact directly with George about calling it from his Kalman filter code and general integration and testing. Please cc all messages to dev(a)dynare.org Many thanks and all the best, Michel andrea pagano wrote: > Hi Michel > > > I was about to send you the codes (which I wrote in fortran). > Actually I used a slight different strategy as one of DGEMM (since I > believe we can specify even more our situation). > > I am finishing to write a couple of extra files and the documentation, > and I will send you the entire package. > > Best > > Andrea > > > > > > > > > > On Tue, May 26, 2009 at 11:35 AM, Michel Juillard > <michel.juillard(a)ens.fr> wrote: > >> Dear Andrea, >> >> were you able to decipher the strategy used by ATLAS for DGEMM? Is there >> hope (and advantages) to specialize it for quasi-triangular matrices? >> >> Best >> >> Michel >> >> > > > >

3 3

Dynare in Ubuntu Karmic
by Sébastien Villemot 21 May '09

21 May '09

It looks like Dynare has now entered Ubuntu, and is in line for the next release in October 2009 (Karmic Koala). https://launchpad.net/ubuntu/+source/dynare This is a consequence of the automatic updates of Ubuntu from Debian. Recall that this first version of the package is an Octave-only version. I still need to add Matlab support (in the form of some build helper). Best, -- Sébastien Villemot

1 1

Dynare in Debian Unstable
by Sébastien Villemot 17 May '09

17 May '09

Hi all, Dynare has entered the official Debian Unstable distribution. At any time you can check the package status on the following page: http://packages.qa.debian.org/d/dynare.html Note that this page has an RSS feed which tracks changes. Best, -- Sébastien Villemot

1 0

parallel dynare
by Marco Ratto 15 May '09

15 May '09

Dear Michel, in order to be able to run the parallel computations under the windows environment, the attached executables must be unzipped and the folder where they are saved must be present in the windows path. I guess we have to find a way to `automatize' this procedure while installing dynare on windows machine? All the best Marco

1 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

Dev May 2009