[CP2K-user] [CP2K:18647] error of NNP simulation

[Vincente Della Balda]

Hi Kordia,

The error message exactly states why the NNP MD does not run: The symmetry 
functions you would like to use are not supported in CP2K (see here at line 
486: https://github.com/cp2k/cp2k/blob/master/src/nnp_environment.F).

If you would like to run the simulation with CP2K you need to stick to the 
radial symmetry function of type 2 and angular one of type 3. Regarding 
cutoff functions, cosine (type 1) and hyperbolic tangent (type 2) are 
supported.

Best,
Vincente

Kordia Bulrat schrieb am Dienstag, 11. April 2023 um 04:02:30 UTC+2:

> Dear Mostafa,
>
> Thank you for your kind suggestion.
> Yes, I trained the potential for the present system. Reference data were 
> obtained from first-principles MD simulation carried out using VASP but 
> smaller cell size.
> I reduced the number of symmetry functions using nnp-prune program in n2p2 
> OK. I will try LAMMPS
>
> Kordia
>
> 2023年4月10日(月) 21:19 Mostafa Abedi <abedimo... at gmail.com>:
>
>> Hi.
>> Did you do the NNP training for this system? Your symmetry function (SF) 
>> set is really big, especially the angular part. Did you optimize the number 
>> of SFs before training?  
>> I haven't tried using NNP in CP2K, but you can have a look at the water 
>> example: 
>>
>> https://github.com/cp2k/cp2k/tree/master/tests/NNP/regtest-1
>> https://github.com/cp2k/cp2k/tree/master/data/NNP/bulkH2O-jcp2020-cnnp
>>
>> I use the LAMMPS interface. You can also try that.
>>
>> Mostafa
>>
>> On Mon, Apr 10, 2023 at 12:36 AM Kordia Bulrat <ryuk... at gmail.com> wrote:
>>
>>> Dear CP2K users,
>>>
>>> I tried to run a molecular dynamics simulation using neural network 
>>> potential (NNP).
>>> I have installed the most recent development version of CP2K and 
>>> prepared necessary files using n2p2 programs. (
>>> https://compphysvienna.github.io/n2p2/index.html)
>>> But the run terminted with an error message "NNP| Symmetry function type 
>>> not supported"
>>> It seems that symmetry functions format in the input.nn file is not 
>>> correct.
>>> I serached the cp2k manual, but I could not find relevant format 
>>> information in cp2k manual.
>>> I attached input, output, and input.nn files.
>>> If anyone has a similar experience,  your suggestion will be of great 
>>> help.
>>>
>>> Thanks in advance, 
>>>
>>> Kordia
>>>
>>> -- 
>>> You received this message because you are subscribed to the Google 
>>> Groups "cp2k" group.
>>> To unsubscribe from this group and stop receiving emails from it, send 
>>> an email to cp2k+uns... at googlegroups.com.
>>> To view this discussion on the web visit 
>>> https://groups.google.com/d/msgid/cp2k/1642fee0-b5ad-4100-b2fd-86ca07b3a93cn%40googlegroups.com 
>>> <https://groups.google.com/d/msgid/cp2k/1642fee0-b5ad-4100-b2fd-86ca07b3a93cn%40googlegroups.com?utm_medium=email&utm_source=footer>
>>> .
>>>
>> -- 
>> You received this message because you are subscribed to the Google Groups 
>> "cp2k" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to cp2k+uns... at googlegroups.com.
>>
> To view this discussion on the web visit 
>> https://groups.google.com/d/msgid/cp2k/CAHQ2YDxbeTRNwd-N8oTe-gXEee3NZ3tQ8PJtgh%2BcE%2Bu7gkv2dQ%40mail.gmail.com 
>> <https://groups.google.com/d/msgid/cp2k/CAHQ2YDxbeTRNwd-N8oTe-gXEee3NZ3tQ8PJtgh%2BcE%2Bu7gkv2dQ%40mail.gmail.com?utm_medium=email&utm_source=footer>
>> .
>>
>

-- 
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+unsubscribe at googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/8925bd42-3b04-492f-aaa0-31b41e051c39n%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20230411/f35a4e92/attachment.htm>