March 2010 - Dev - dynare.org

[Trac] #91: Univariate smoother with missing observations needs corrections.
by Trac Dynare 03 Apr '10

03 Apr '10

#91: Univariate smoother with missing observations needs corrections. --------------------------+------------------------------------------------- Reporter: stepan | Owner: Type: bug | Status: new Priority: major | Milestone: Component: Core M-files | Version: 4.1.0 Keywords: | --------------------------+------------------------------------------------- The last block (if nargout>7) is wrong and cause dynare to crash. Variable iF is not defined in this function (should be Fi i think). There is also a problem with the selection matrix Z. -- Ticket URL: <https://www.dynare.org/trac/ticket/91> Dynare <http://www.dynare.org> The Dynare project

1 2

[Trac] #74: Variance decomposition gives dummy results in the presence of non-stationary vars
by Trac Dynare 03 Apr '10

03 Apr '10

#74: Variance decomposition gives dummy results in the presence of non- stationary vars --------------------------+------------------------------------------------- Reporter: sebastien | Owner: Type: bug | Status: new Priority: major | Milestone: Component: Core M-files | Version: 4.1.0 Keywords: | --------------------------+------------------------------------------------- In the attached MOD file, the variance decomposition is clearly bogus: some lines don't sum to 100% (some percentages are even negative). The problem does not occur with Dynare 4.0.4. This is probably related to the fact that this model contains a non-stationary variable. The file {{{th_autocovariances.m}}} has changed between versions 4.0.4 and 4.1.0 with respect to the treatment of stationary variables. Also note that the problem does not come from the fact that some shocks are switched off (with a zero variance) in the attached MOD file. Activating all the shocks still gives dummy results. -- Ticket URL: <https://www.dynare.org/trac/ticket/74> Dynare <http://www.dynare.org> The Dynare project

1 2

Re: [DynareDev] [Trac] #21: Add partial information
by Trac Dynare 24 Mar '10

24 Mar '10

#21: Add partial information ---------------------------+------------------------------------------------ Reporter: sebastien | Owner: george Type: feature | Status: accepted Priority: major | Milestone: 4.2 Component: Core M-files | Version: Resolution: | Keywords: ---------------------------+------------------------------------------------ Changes (by sebastien): * milestone: => 4.2 Comment: As of 2010-03-24, part of the work has been done. Two minor issues to be solved (among others): * the file matlab/disclyap_fast.m has no copyright: this should be clarified * the files matlab/partial_information/{qzswitch,qzdiv}.m are redundant with the versions in matlab/qz. These files should be deleted once the PI code relies on mjdgges. -- Ticket URL: <https://www.dynare.org/trac/ticket/21#comment:3> Dynare <http://www.dynare.org> The Dynare project

1 0

options_.partial_information
by George Perendia 19 Mar '10

19 Mar '10

Hi Michel and Sebastien (et al) The partial information is now failing in estimation_1.m (on line 1096) since there is no PI Kalman smother available yet. As an interim solution we suggest to bypass that whole if ... diagnostics block by adding an additional clause & options_.partial_information == 0 to its if test on lines 1086-1088 , i.e. if (~((any(bayestopt_.pshape > 0) & options_.mh_replic) | (any(bayestopt_.pshape ... > 0) & options_.load_mh_file)) ... | ~options_.smoother ) & M_.endo_nbr^2*gend < 1e7 ... & options_.partial_information == 0 % to be fixed . Would that be ok? -- Best regards George

2 1

[Trac] #82: prior_analysis and posterior_analysis commands broken
by Trac Dynare 18 Mar '10

18 Mar '10

#82: prior_analysis and posterior_analysis commands broken --------------------------+------------------------------------------------- Reporter: sebastien | Owner: sebastien Type: bug | Status: new Priority: major | Milestone: Component: Preprocessor | Version: 4.1.0 Keywords: | --------------------------+------------------------------------------------- The prior_analysis and posterior_analysis commands are broken: the calling sequence generated by the preprocessor doesn't match the calling sequence of the MATLAB function... Their documentation in the ref. manual also needs updating. -- Ticket URL: <https://www.dynare.org/trac/ticket/82> Dynare <http://www.dynare.org> The Dynare project

1 1

[Trac] #65: Second derivatives w.r.t. parameters
by Trac Dynare 15 Mar '10

15 Mar '10

#65: Second derivatives w.r.t. parameters --------------------------+------------------------------------------------- Reporter: ratto | Owner: sebastien Type: feature | Status: new Priority: major | Milestone: 4.2-alpha2 Component: Preprocessor | Version: Keywords: | --------------------------+------------------------------------------------- In identification analysis, for the computation of the Jacobian of typical optimization problems like moment matching (limited information). -- Ticket URL: <https://www.dynare.org/trac/ticket/65> Dynare <http://www.dynare.org> The Dynare project

1 5

Re: [DynareDev] [Trac] #19: Add tool to automatically write a stationary version of the model, given trends supplied by user
by Trac Dynare 12 Mar '10

12 Mar '10

#19: Add tool to automatically write a stationary version of the model, given trends supplied by user ---------------------------+------------------------------------------------ Reporter: sebastien | Owner: houtanb Type: feature | Status: accepted Priority: major | Milestone: 4.2 Component: Preprocessor | Version: Resolution: | Keywords: ---------------------------+------------------------------------------------ Changes (by houtanb): * owner: sebastien => houtanb * status: new => accepted -- Ticket URL: <https://www.dynare.org/trac/ticket/19#comment:3> Dynare <http://www.dynare.org> The Dynare project

1 0

Re: [DynareDev] Estimation DLL design
by George Perendia 12 Mar '10

12 Mar '10

Hi If nspred is number of predetermined variables in the state vector defined in set_state_space.m as dr.nspred = sum(kstate(:,2) <= M_.maximum_endo_lag+1); Can nspred then be larger than npred, i.e. the total number of predetermined variables (=nboth + only predetermined)? , i.e.in dynare_estimation_1.m we have: bayestopt_.restrict_columns = [ic; length(k2)+(1:nspred-npred)']; but the differnce ia then not supposed to be larger than 0 and the second term in the square brackets should then never yield anything other than zero length vector so it appears redundant - or am I missing something. -- Best regards George

1 0

[Trac] #32: Report MATLAB crashes on Linux/64 to Mathworks
by Trac Dynare 10 Mar '10

10 Mar '10

#32: Report MATLAB crashes on Linux/64 to Mathworks --------------------------+------------------------------------------------- Reporter: sebastien | Owner: Type: bug | Status: new Priority: major | Milestone: Component: Core M-files | Version: Keywords: | --------------------------+------------------------------------------------- On Linux, the 64-bits version of MATLAB crashes on some models with Dynare. The crash seems to occur with rather big models. It is reproducible, in the sense that the same code will always trigger the bug; but a small change in the code can sometimes make the crash disappear. Note that the crash is only triggered by M-files (no MEX). See the attached "plouf.m" and "plouf_static.m". They are derived from preprocessor-generated files with the EAGLE model. The script "plouf.m" calls the function "plouf_static.m". Attached is the crash dump of MATLAB. The crash is reproducible on MATLAB Linux/64, versions 7.5, 7.6 and 7.8. But it does not occur on MATLAB for Linux/32, Windows/32 and Windows/64. As can be seen from the displayed marks which I have inserted, MATLAB crashes around line 131 of "plouf_static.m". More precisely, it seems to crash just after having computed the expression there. I can't understand exactly what's going on. We need to report this to Mathworks. -- Ticket URL: <https://www.dynare.org/trac/ticket/32> Dynare <http://www.dynare.org> The Dynare project

1 3

[Trac] #60: Fix memory leaks in k_order_perturbation DLL
by Trac Dynare 09 Mar '10

09 Mar '10

#60: Fix memory leaks in k_order_perturbation DLL -------------------------+-------------------------------------------------- Reporter: sebastien | Owner: george Type: bug | Status: new Priority: minor | Milestone: Component: K-order DLL | Version: Keywords: | -------------------------+-------------------------------------------------- -- Ticket URL: <https://www.dynare.org/trac/ticket/60> Dynare <http://www.dynare.org> The Dynare project

1 6