Thanks for the reminder. :-)

> Including support for Quantum Espresso with Avogadro is not quite an 
> idealistic thing to do at the moment, as it will have the same problem as 
> faced by VASP, i.e. apart from the positions of the various atomic species, 
> one also needs to provide a couple of other details like energy cutoffs, 
> pesudopotentials etc in the same input file, unlike other popular packages 
> where there is a standalone file for atomic positions.

Agreed - let's have a separate discussion on the mailing list or ideally on 
http://discuss.avogadro.cc about this. I think there are a number of VASP / 
Quantum Espresso users and I'd certainly like to see some support for these 
packages.

> Meanwhile, I have been fixing minor bugs I have been coming across in 
> Avogadro, and that has also given me the chance to look and examine through 
> various scripts. And since I have about two weeks before the proposal 
> deadline, I would like to get more information on the plan to include support 
> for MD. 
>       • I guess we would be starting with parsing the output files of the MD 
> calculations and getting details of each time step. First ideas that struck 
> my mind were along the lines of including a slider for navigating to the n'th 
> timestep (or playing the time step evolution as an animation), so that one 
> can visualize how the atoms move along each time step.

Right. We'd want a timeline with a slider and/or the ability to jump to 
particular spots. For example, if an important transition occurs at the n'th 
step (e.g., 305 ps) you should be able to put in that time value and jump.

This would certainly include improvements on the current "player" tool.

>       • Focusing on the reading of large files is an interesting aspect to 
> explore, both algorithmically and through code. I shall read more about some 
> of the common encoding techniques and try to investigate whether techniques 
> for time series analysis like data differencing, can help get better insights 
> on the data we have.

Many of the MD trajectories include compressed formats, but as we mentioned 
previously, I think this is a good task for the project - can modern 
compression techniques offer something over the current MD formats?

>       • Once the major backbone is dealt with, focusing on 
> cross-compatibility with various MD codes is to be explored. Whether 
> following JSON setup similar to the one already existing for the core 
> framework, or adopting some other writing techniques, is to be explored too.

I think another key aspect would be to consider ways of exporting 
visualizations, e.g. programatic export of PNG images would be an easy first 
step, since there's already an export graphics method. 

I feel that it would be good to have some analysis tools. For example, 
generating radial distribution plots, etc. How does the energy change over 
time? Does the volume change over the timeline? Can you plot a particular angle 
or distance across the timeline?
(One might even want to have a set of "channels" for monitoring different 
properties as strip charts across the trajectory.)

There are a few packages that could either be adapted or at least point to 
particular key tasks:
* MD Analysis: https://www.mdanalysis.org/about/
* MDTraj: http://mdtraj.org/1.9.0/
* VMD plugins: http://www.ks.uiuc.edu/Research/vmd/plugins/

> Also, I would like to know whether you are comfortable with having a look at 
> proposal drafts before the official submission deadline.

Yes, absolutely. We generally provide feedback on drafts. We like to see the 
best proposals possible. Please feel free to post a draft and we can give some 
comments.

Hope that helps,
-Geoff
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Avogadro-devel mailing list
Avogadro-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/avogadro-devel

Reply via email to