[ase-users] no energies in .traj file when running neb calculations with siesta calculator

[Louwerse, M.J. (Manuel)]

Thanks.
I did have to change "label='image_{:02d}'.format(i)" to "label='image_'+str(i)" to get it working, but I guess that is a python-version issue.
Also in your script there is no calculator connected to the initial and final images, so the get_potential_energy lines don't work for them...

But after correcting that, it works perfect now, and you also solved another issue I had with the efficiency, since each image was restarting from the previous image instead of from the previous step of the same image... Perhaps the trick to label the images with a number should also be mentioned in the manuals on the ase website.

Manuel


>There was a typo in my previous script. I've attached the fixed version (name -> fname).

>Eric

On Mon, 2017-04-10 at 15:41 +0000, Eric Hermes via ase-users wrote:
> On Mon, 2017-04-10 at 14:56 +0000, Louwerse,	M.J. (Manuel) via
> ase-
> users wrote:
> > Dear All,
> >  
> > When I am running nudged elastic band (neb) calculations with Siesta 
> > as calculator, my trajectory files do not contain any energies.
> > When I run neb calculations with the simple EMT calculator I do get 
> > energies written to my trajectory file, and when I do a simple 
> > minimization with the siesta calculator also. But when using the 
> > combination of neb and siesta no energies are written.
> > Does anyone have any idea what may cause this and how it can be 
> > solved?
> >  
> > Thanks,
> > Manuel
> 
> Manuel,
> 
> I thought I had replied to your previous email, but I don't see my 
> reply, so it looks like I didn't.
> 
> Please see the attached script, which is modified from the example you 
> sent in your previous email. I have made some changes which are merely 
> semantic (i.e. we would normally say that your example has 8 images, 
> not 10, since we do not perform calculations on either the initial or 
> final images), but others are functional.
> 
> For example, each image should get its *own* calculator object. While 
> your code may work as written for serial calculations, it is very 
> strange to move a calculator from Atoms object to Atoms object. If 
> each image gets its own calculator, you will also be able to run the 
> NEB in parallel (that is, with each image running simultaneously on 
> different processors by calling NEB with parallel=True).
> 
> I also am not sure what the final few lines of your script was trying 
> to do. As written, your script will run the NEB and create a 
> trajectory file where each group of 10 sequential images is a single 
> iteration of the NEB, but then you appear to overwrite it with an 
> empty file. I think what you are trying to do is read the trajectory 
> file written by NEB, but you need to do that with ase.io.read, not 
> ase.io.trajectory.
> 
> Let me know if this doesn't work with your system. I do not have 
> experience with Siesta or the Siesta calculator in ASE, so there might 
> be issues I am unaware of.
> 
> Eric
> > _______________________________________________
> > ase-users mailing list
> > ase-users at listserv.fysik.dtu.dk
> > https://listserv.fysik.dtu.dk/mailman/listinfo/ase-users
> 
> _______________________________________________
> ase-users mailing list
> ase-users at listserv.fysik.dtu.dk
> https://listserv.fysik.dtu.dk/mailman/listinfo/ase-users