Octave 3.8.1 and Dynare

[jmgb]

Hello Dynare team,

yesterday I decided to install Dynare for Octave on my Ubuntu 12.04 LTS system.

First step:
I installed octave 3.8.1 (latest stable version has now a GUI frontend) via Octave's Personal Package Archive (PPA).
LINK: http://wiki.octave.org/Debian
No problems there.

Second step:
I followed your clearly written instructions on this site to install Dynare:
http://www.dynare.org/DynareWiki/Instal ... anOrUbuntu
Problem here: When I try to install dynare 4.4.2 the Synaptic package manager wants to deinstall
octave 3.8 simultaneously. (I guess to install Dynare side by side with the old octave 3.2 version.)
I tried the following APT source lines (with no luck):
deb http://www.dynare.org/ubuntu precise main contrib [for octave 3.6(from Octave PPA)]
deb http://www.dynare.org/ubuntu trusty main contrib [for octave 3.8]

Is there any possibility for Ubuntu 12.04 LTS users, who updated to octave 3.8.1 via PPA, to get access to a version of Dynare?

[SébastienVillemot]

My understanding of the problem is the following: the Dynare package for 12.04 (precise) has been compiled with Octave 3.6.

Then the Octave PPA was updated to octave 3.8. But we did not update our package, so you are now stuck.

I'll try to update our package soon. I'll keep you informed.

[excessliquid]

I thought Dynare began compatibility with Octave 3.8 in version 4.4.1 onwards?
I also have a problem using Dynare with octave 3.8 - when running a .mod that works in octave 3.6 I get an error saying "could not find library or dependents: C:\dynare\4.4.2\matlab\../mex/octave\mjdgges.mex".

Oddly though at the start dynare seems to confirm all the .mex files are available:
"Configuring Dynare ...
[mex] Generalized QZ.
[mex] Sylvester equation solution.
[mex] Kronecker products.
[mex] Sparse kronecker products.
[mex] Local state space iteration (second order).
[mex] Bytecode evaluation.
[mex] k-order perturbation solver.
[mex] k-order solution simulation.
[mex] Quasi Monte-Carlo sequence (Sobol).
[mex] Markov Switching SBVAR."

[SébastienVillemot]

Yes Octave 3.8 compatibility started at Dynare 4.4.1, but at the source code level. Binaries need to be updated, and we did not update our Ubuntu Precise binaries for Octave 3.8.

[SébastienVillemot]

I have now updated the binaries for Ubuntu Precise. They should work with Octave 3.8. Please upgrade, and report any problem.

[excessliquid]

I am actually using windows. Octave 3.6 has a bug which causes it to crash when producing too many plots (ft_render error) so upgraded to 3.8.

Would a similar fix be available for the windows installation?

[jmgb]

Dear Sébastien Villemot,
thank you very much for your efforts to make the newest Dynare version available for Ubuntu 12.04 LTS and Octave 3.8.1 ppa users.
I downloaded the newest binaries for Dynare (that means dynare, dynare-doc,dynare-matlab, dynare-common, all of them in the 4.4.2.1~precise2 version) with Synaptic. (I had to restrict the package sources to the Dynare ATP lines. Before I did that, I got only the older 4.2.1 version [I guess from the the ubuntu repository]). Now, when I mark dynare 4.4.2.1~precise2 for installation I get a warning that the dynare package has a defect. There is an exclamation mark before it.

[SébastienVillemot]

Unfortunately, without more information about the nature of the defect that synaptics has found, I can hardly help… On my side the package looks fine.

Maybe you did not import the cryptographic key that is used to sign the custom Dynare packages?

[jonessssss]

For some reason (like a shocking memory) I forgot about my Dynare in sourceslist and downloaded and compiled from the sources. Mostly it seems ok. It is good to look at the configure output and install as many of the related packages as possible as this speeds up the compile time a hell of a lot. Check your dkms which I think definitely should be loaded.

Ubuntu 12.04.
I had got myself into a snag with Rstudio Version 0.98.501 and R version 3.1.1 (2014-07-10) -- "Sock it to Me"
Copyright (C) 2014 The R Foundation for Statistical Computing
Platform: i686-pc-linux-gnu (32-bit).

If I installed R it would remove octave libblas, libAtlas and one other linear algebra library or vice versa.

I don't know perhaps it was the upgrade that caused the incompatibility (not a dynare problem) but I believe I have both running Ok now and R packages compiling and not failing with can't find lblas.

I suspect it is my ignorance in not knowing how to use update-alternatives.

I am sorry this is not much help but it does seem to compile and install ok from source. I will try and work up the list of packages I added to get everything running.

[luisjaime]

Good day for all:

I'm a linux user and I've this problem:

I've installed octave 3.8.1 from ppa, I've downloaded dynare 4.4.3, I did configure and make, and make install, all without errors but octave don't recognize the command dynare, always say: "'dynare' undefined near line 1 column 1". What I can do for fix this problem?

Thanks,

luis jaime

[SébastienVillemot]

luisjame: why don't you simply use the Dynare pre-compiled package that we offer?

[dmbn]

SébastienVillemot, can you say, whether the building up of rpm-packages is scheduled? Because I use, for example, an rpm-based distribution, and compiling from source files make many difficulties... (see http://www.dynare.org/phpBB3/viewtopic.php?f=1&t=5709)

[SébastienVillemot]

@dmbn: as far as I know, this is not currently planned. It is a lot of work to support many distributions. Any help is of course welcome.