[ADMB Users] [Developers] FW: myxomatosis example

bigA jaguar.smart at gmail.com
Tue Aug 2 21:43:10 PDT 2011


Just out of curiosity, why not have a second package for user contributed
code? Then as individual users want to extend their personal library of admb
functions they would already have a standardized build process with a make
file already started to build from.   It could also help establish some
coding conventions,  across different user's contributed code, and
a coherent way for users to include documentation  with the functions they
submit. Finally it could also be useful to ensure that credit is given to
those who contribute; admb is an open source project and it is nice make
sure that people's names are on their work.  I would be willing to help with
this, but I want to know how the admb establishment feels.
Aaron

On Tue, Aug 2, 2011 at 7:58 PM, Steve Martell <s.martell at fisheries.ubc.ca>wrote:

> I think Johnoel's idea of a contrib.lib and contrib.h would work best, and
> the end user would still have to have put #include <contrib.h> in the
> GLOBALS_SECTION.  I'm hesitant to modify the existing makefiles to have the
> buildbot include the contributed code in the builds, so I did not do this.
>  It should be possible, however, to create a makefile in the contrib
> directory that would build the contrib.lib and put this in the lib directory
> of the build.  Makefiles are pretty new to me so I'd need some help on this
> one.
>
> Steve
> On 2011-08-02, at 2:02 PM, Johnoel Ancheta wrote:
>
> > I may have a easy solution, but will need to test.  It involves creating
> a contrib.lib and contrib.h that
> > gets built with admb.  Easier for the end user...
> >
> > On 8/2/11 10:48 AM, Mark Maunder wrote:
> >> John,
> >>
> >> I thought that the idea was to have the contributed code compiled in the
> libraries, but not include the #include statement somewhere so that the user
> had to physically put the include statement in their tpl (similar to how an
> R user has to physically load the library). The messy part is having to use
> the whole path to the code and not just the name of the library
> >>
> >> Mark
> >>
> >> -----Original Message-----
> >> From: developers-bounces at admb-project.org [mailto:
> developers-bounces at admb-project.org] On Behalf Of John Sibert
> >> Sent: Tuesday, August 02, 2011 1:41 PM
> >> To: developers at admb-project.org
> >> Subject: Re: [Developers] FW: myxomatosis example
> >>
> >> Well including stuff in the GLOBALS_SECTION does not seem particularly
> messy, Dave is correct the usual way is with a header file and a library.
> >>
> >> The practical question is how are the libraries assembled? Each
> OS/architecture/compiler combination brings unique problems. That is why we
> have the buildbot for the ADMB code. It might be feasible to include the
> contributed contribted code, but in my opinion the ADMB Project should not
> be responsible for maintenance of user-contributed code.
> >> Could we ask code contributors to create an appropriate header file and
> a makefile to create a library for their code? If so then users of the
> contributed code would have the option of either including the source code
> in the tpl or customizing the makefile to build libraries appropriate to
> their system.
> >>
> >> John
> >>
> >> On 08/01/2011 02:08 PM, dave fournier wrote:
> >>> On 11-08-01 04:15 PM, Mark Maunder wrote:
> >>>
> >>> The usual way is to include a header file and then to link in a
> >>> library with the compiled code.
> >>>
> >>>> Can anyone give advice on how to include user contributed libraries
> >>>> in the Globals section. This looks too messy. Can't you just do
> >>>> "#include ecolib"
> >>>>
> >>>> Thanks,
> >>>>
> >>>> Mark
> >>>>
> >>>> *From:*Mollie Brooks [mailto:mbrooks at ufl.edu]
> >>>> *Sent:* Monday, August 01, 2011 4:03 PM
> >>>> *To:* Mark Maunder
> >>>> *Subject:* myxomatosis example
> >>>>
> >>>> Hi Mark,
> >>>>
> >>>> I did an example from Ben's book. I'm also working on another one.
> >>>> This one requires a vectorized version of the dgamma function which I
> >>>> just added to Steve's statlib on the svn. I'd like any suggestions
> >>>> you might have, especially on the globals section. I think the
> >>>> developers should have a talk about making it as easy as possible for
> >>>> users to include libraries from the contrib directory. Maybe these
> >>>> should get moved to the include directory automatically when we
> >>>> compile. Also, the library pow_vectorized_RE should probably be
> >>>> incorporated into the core source.
> >>>>
> >>>> I did it using R2admb which currently doesn't work for windows, but
> >>>> I'm hoping Ben will have it working later this month.
> >>>>
> >>>> cheers,
> >>>>
> >>>> Mollie
> >>>>
> >>>>
> >>>> _______________________________________________
> >>>> Developers mailing list
> >>>> Developers at admb-project.org
> >>>> http://lists.admb-project.org/mailman/listinfo/developers
> >>>
> >>> _______________________________________________
> >>> Developers mailing list
> >>> Developers at admb-project.org
> >>> http://lists.admb-project.org/mailman/listinfo/developers
> >> --
> >> John Sibert
> >> Emeritus Researcher, SOEST
> >> University of Hawaii at Manoa
> >>
> >> Visit the ADMB project http://admb-project.org/
> >>
> >>
> >> _______________________________________________
> >> Developers mailing list
> >> Developers at admb-project.org
> >> http://lists.admb-project.org/mailman/listinfo/developers
> >> _______________________________________________
> >> Developers mailing list
> >> Developers at admb-project.org
> >> http://lists.admb-project.org/mailman/listinfo/developers
> >
> > _______________________________________________
> > Users mailing list
> > Users at admb-project.org
> > http://lists.admb-project.org/mailman/listinfo/users
>
> _______________________________________________
> Users mailing list
> Users at admb-project.org
> http://lists.admb-project.org/mailman/listinfo/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.admb-project.org/pipermail/users/attachments/20110802/f7da21a3/attachment.html>


More information about the Users mailing list