[ase-users] ***SPAM*** Re: info NEB+Siesta
Ask Hjorth Larsen
asklarsen at gmail.com
Sat Jul 1 18:49:40 CEST 2017
Oh, now I see the problem: If the images have the same calculator, the
energies are not saved. This behaviour seems very erratic so I will
open an issue and discuss how the case should be handled.
Workaround: Create distinct calculators for each image.
http://dcwww.camd.dtu.dk/~askhl/files/gpaw-neb-one-calc-per-image.py
Best regards
Ask
2017-06-30 21:35 GMT+02:00 marc <marc.barbry at mailoo.org>:
> Hi Ask,
>
> I don't see any energy store in the traj file.
> $ ase info A2B_gpaw.traj -v
> A2B_gpaw.traj: ASE trajectory (traj)
> A2B_gpaw.traj (tag: "ASE-Trajectory", 5 items)
> item #0:
> {
> ase_version: 3.14.0b1,
> cell: [[ 30. 0. 0.]
> [ 0. 30. 0.]
> [ 0. 0. 30.]],
> description: {u'type': u'optimization', u'interval': 1, u'optimizer':
> u'MDMin'},
> numbers: <ndarray shape=(8,) dtype=int64>,
> pbc: [False False False],
> positions: <ndarray shape=(8, 3) dtype=float64>,
> version: 1}
>
> As you can see I'm using 3.14.0b1.
>
> I will update and try again.
>
> Best regards,
> Marc
>
>
>
> On 06/30/2017 06:05 PM, Ask Hjorth Larsen wrote:
>>
>> Hi Marc
>>
>> I cannot reproduce the issue. It appears to me that energies are
>> correctly stored in trajectories when running NEB with GPAW. Which
>> version of ASE are you using?
>>
>> Best regards
>> Ask
>>
>> 2017-06-30 12:32 GMT+02:00 marc <marc.barbry at mailoo.org>:
>>>
>>> Hi Ask,
>>>
>>> I attached the input file than Tommy sends and the output A2B_siesta.traj
>>> and A2B_gpaw.traj. The gui is giving in the both case similar results.
>>> The
>>> problem is probably at the writing of the files.
>>>
>>> Best regards,
>>> Marc
>>>
>>> On 06/29/2017 05:17 PM, Ask Hjorth Larsen wrote:
>>>>
>>>> Actually wait - if it affects GPAW too, it must be something else.
>>>> Might it be because the end points are part of the trajectory but do
>>>> not have energies? Anyway running the same command and posting the
>>>> output would help.
>>>>
>>>> Best regards
>>>> Ask
>>>>
>>>> 2017-06-29 17:10 GMT+02:00 Ask Hjorth Larsen <asklarsen at gmail.com>:
>>>>>
>>>>> Hi
>>>>>
>>>>> 2017-06-28 14:03 GMT+02:00 marc via ase-users
>>>>> <ase-users at listserv.fysik.dtu.dk>:
>>>>>>
>>>>>> It may be a bug in the siesta calculator. its seems that the different
>>>>>> field
>>>>>> necessary for neb calculations are not written in the traj file. I
>>>>>> will
>>>>>> have
>>>>>> a look.
>>>>>>
>>>>>> Maybe someone that knows well neb calculations could help me with
>>>>>> this?
>>>>>>
>>>>>> Marc
>>>>>
>>>>> There could be some irregularity of how the Siesta calculator handles
>>>>> its results, so energies are not recorded in trajectories. I seem to
>>>>> remember something related happening. Can someone who has Siesta
>>>>> installed perform a quick relaxation of a small molecule and check (or
>>>>> send) the trajectory file?
>>>>>
>>>>> e.g.: ase info mytrajectoryfile.traj -v
>>>>>
>>>>> And also check the NEB trajectories similarly, in case there's a
>>>>> difference.
>>>>>
>>>>> The question is whether the "energy" entry is there.
>>>>>
>>>>> Here's an excerpt from one of my trajectry files:
>>>>>
>>>>> askhl at jormungandr:~$ ase info opt.traj -v
>>>>> opt.traj: ASE trajectory (traj)
>>>>> opt.traj (tag: "ASE-Trajectory", 3 items)
>>>>> item #0:
>>>>> {
>>>>> ase_version: 3.14.0b1,
>>>>> calculator: {
>>>>> dipole: [ -1.08619140e-13 -2.88470856e-16 8.72627641e-14],
>>>>> energy: -6.74028940919,
>>>>> forces: <ndarray shape=(2, 3) dtype=float64>,
>>>>> name: gpaw,
>>>>> parameters: {u'xc': {u'name': u'vdW-DF-CX', u'libvdwxc_name':
>>>>> u'vdW-DF-CX', u'vdwcoef': 1.0, u'setup_name': u'revPBE', u'pfft_grid':
>>>>> None, u'mode': u'auto', u'type': u'libvdwxc', u'semilocal_xc':
>>>>> u'GGA_X_LV_RPW86+LDA_C_PW'}, u'mode': u'pw'}},
>>>>> cell: [[ 4.1 0. 0. ]
>>>>> [ 0. 4. 0. ]
>>>>> [ 0. 0. 4.737166]],
>>>>> description: {u'type': u'optimization', u'maxstep': 0.04,
>>>>> u'optimizer': u'BFGS', u'interval': 1},
>>>>> numbers: <ndarray shape=(2,) dtype=int64>,
>>>>> pbc: [False False False],
>>>>> positions: <ndarray shape=(2, 3) dtype=float64>,
>>>>> version: 1}
>>>>> (....)
>>>>>
>>>>> Best regards
>>>>> Ask
>>>>>
>>>>>> On 06/28/2017 01:25 PM, Tommaso Francese wrote:
>>>>>>
>>>>>> Hi Marc!
>>>>>> First of all thanks a lot for testing my system.
>>>>>>
>>>>>> The movie is only a part of what I would like to have in the end of
>>>>>> the
>>>>>> day.
>>>>>> The most important thing it would be the obtaining of the energy
>>>>>> profile
>>>>>> computed by siesta through the Neb module. In simple terms, I would
>>>>>> like
>>>>>> to
>>>>>> visualize with the ASE-gui/NEB utility the energy profile, see this
>>>>>> example
>>>>>> for clarity:
>>>>>>
>>>>>> https://wiki.fysik.dtu.dk/ase/tutorials/neb/diffusion.html
>>>>>>
>>>>>> The point is the examples works fine, with the EMT algorithm, that it
>>>>>> implemented into the ASE, but it looks like to fail while coupled with
>>>>>> Siesta, in this case. It looks like it is not able to read the
>>>>>> energies,
>>>>>> or
>>>>>> to extract them from the final file generated.
>>>>>>
>>>>>> Thanks for the help,
>>>>>> Best,
>>>>>> Tommy
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Il giorno 28 giu 2017, alle ore 13:13, marc via ase-users
>>>>>> <ase-users at listserv.fysik.dtu.dk> ha scritto:
>>>>>>
>>>>>> Hello Tommy,
>>>>>>
>>>>>> I have never used NEB calculation before so I'm not sure what you want
>>>>>> to
>>>>>> get.
>>>>>> I started to run the calculations and the neb.traj file is written.
>>>>>> Opening
>>>>>> it with the ase -T gui command gives me a movie with the molecule
>>>>>> moving. Is
>>>>>> it what you want to get?
>>>>>>
>>>>>> Best,
>>>>>> Marc
>>>>>>
>>>>>> On 06/28/2017 11:16 AM, Tommaso Francese via ase-users wrote:
>>>>>>
>>>>>> Dear all Asusers,
>>>>>>
>>>>>> I am pretty newbie in the Ase world. I was trying to run a NEB
>>>>>> calculation
>>>>>> making use of the Siesta code. I made use of a template for the Orca
>>>>>> code,
>>>>>> properly modified. The code runs fine, but when I try to visualize the
>>>>>> output via ASE-gui with the NEB module, nothing is reported.
>>>>>>
>>>>>> I don’t know where the error lies, can you kindly help me fix this
>>>>>> problem?
>>>>>>
>>>>>> Here I provide the script for the running the NEB coupled with Siesta,
>>>>>> and
>>>>>> the two input files (react.traj and prod.traj) that are interpolated
>>>>>> for
>>>>>> creating the NEB path. They latter contains the toy model I am
>>>>>> testing.
>>>>>>
>>>>>> Thanks a lot for the help,
>>>>>> It ’s extremely important.
>>>>>>
>>>>>> Best,
>>>>>> Tommy
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Here also the pseudos for the Siesta (they are in testing mode):
>>>>>>
>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> 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
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> ase-users mailing list
>>>>>> ase-users at listserv.fysik.dtu.dk
>>>>>> https://listserv.fysik.dtu.dk/mailman/listinfo/ase-users
>>>
>>>
>
More information about the ase-users
mailing list