[CP2K-user] [CP2K:19225] Re: The force output of RETRAJ module under a r2scan functional is zero

[Tianhua Wang]

According to my tests, this bug exists in almost all CP2K versions later 
than 6.1, i.e., 7.1, 8.1, 8.2, 9.1, 2022.1, and 2023.2 (I did not test 
2023.1), for both SCAN and r2SCAN functionals. It seems to be a systemic 
problem.

在2023年9月7日星期四 UTC+8 11:20:17<qin li> 写道：

> The force output of the RETRAJ module works problematiclly in version 
> cp2k2022.1.
>
> qin li <flytomoo... at gmail.com> 于2023年9月7日周四 10:20写道：
>
>> Hi,
>>
>> I'm trying to train a deep potential model of the MgCl2 solution based on 
>> r2scan functional, where the positions and forces are needed. I employed 
>> the r2scan functional to rerun the trajectory resulting from the BLYPD3 
>> functional, but the force output of the RETRAJ module comes out to be all 
>> zero. However, the RETRAJ module works fine for the BLYPD3 functional. Does 
>> anyone have an idea of how to resolve this?
>>
>> The cp2k version is cp2k2022.1, inputs of the r2scan and force output are 
>> as follows and the attachment:
>> &XC
>>   &XC_FUNCTIONAL
>>     &MGGA_X_R2SCAN
>>     &END MGGA_X_R2SCAN
>>     &MGGA_C_R2SCAN
>>     &END MGGA_C_R2SCAN
>>   &END XC_FUNCTIONAL
>>   &XC_GRID
>>     XC_DERIV SPLINE3
>>     XC_SMOOTH_RHO NONE
>>   &END XC_GRID
>> &END XC
>>
>> &REFTRAJ
>>   EVAL_ENERGY_FORCES
>>   TRAJ_FILE_NAME ${REFTRAJ}
>> &END REFTRAJ
>>
>> &PRINT
>>   &FORCES
>>     &EACH
>>       MD 1
>>     &END EACH
>>   &END FORCES
>> &END PRINT
>>
>>

-- 
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/5e53f286-fc05-4a28-82c9-3c6e666d49bdn%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20230907/680d45f7/attachment-0001.htm>