[gmx-developers] GMX library functions and documentation

[Ondrej Marsalek]

Thanks a lot for the information. I will work on the tool some more in
the old style and convert it when the new framework is closer to
production. I am certainly looking forward to it, especially the
selection language.

Ondrej

On Fri, Sep 14, 2012 at 6:17 PM, Roland Schulz <roland at utk.edu> wrote:
>
>
> On Fri, Sep 14, 2012 at 11:58 AM, Ondrej Marsalek
> <ondrej.marsalek at gmail.com> wrote:
>>
>> On Fri, Sep 14, 2012 at 5:02 PM, Roland Schulz <roland at utk.edu> wrote:
>> >
>> > The Gromacs 5 analysis tool framework is already documented very well
>> >
>> > (http://jenkins.gromacs.org/job/Doxygen_Nightly_master/javadoc/html-user/group__module__trajectoryanalysis.html)
>> > and the parts I'm using work very well. I personally would write new
>> > analysis tools for Gromacs 5. If you want to try it out, you can get it
>> > from
>> > git.
>>
>> I would very much like to do that, but I would also like to make the
>> tool available to other users (of stable GMX) relatively soon. Is the
>> new framework only getting released in 5.0, or already in 4.6?
>
> It will not be release as part of 4.6. 5.0 might be release Aug 2013. See
> http://redmine.gromacs.org/issues/949 for the preliminary planning.
>
>>
>> Are
>> there any tools already available (like a rewrite of g_rdf or
>> something similar) that I could read?
>
> g_angel and g_select is available. You can run them with "g_ana
> select(/angle)".
>
>>
>> I have seen the template, but a
>> full tool would be nice. Also, what about support for time-dependent
>> quantities, like MSD or time correlation functions?
>
> Any computation which has a 1-1 mapping between the trajectory and the
> output data doesn't require anything special. This should be true also for
> MSD.
> The need for extra computation which needs to happen after the per frame
> computation is build in the concept. But only some functions (e.g. averages,
> distances,  and histograms see
> http://jenkins.gromacs.org/job/Doxygen_Nightly_master/javadoc/html-user/group__module__analysisdata.html)
> are implemented so far. It should be very easy to add a AnalysisDataModule
> which computes correlation  functions (and it is certainly planned).
>
> Roland