[CP2K-user] accessing qs_envs in a multiple force_eval run

[Vladimir Rybkin]

Dear Pavan,

probably, the right place is in force_env_methods.F. There the separate 
force evaluations are called.

Yours,

Vladimir

понедельник, 3 декабря 2018 г., 10:31:13 UTC+1 пользователь Pavan Kumar 
Behara написал:
>
> Hello CP2K developers,
>
> I want to pass the qs_env of two separate force evaluations in a multiple 
> force_eval calculation to another subroutine where I will do some post 
> analysis on the molecular orbitals. I am placing a function call inside 
> force_env_methods.F to define the two states qs_env_a and qs_env_b, and 
> passing them to the other subroutine as: 
>
> CALL force_env_get(force_env=force_env%sub_force_env(1)%force_env, 
> qs_env=qs_env_a)
> CALL force_env_get(force_env=force_env%sub_force_env(2)%force_env, 
> qs_env=qs_env_b)
> CALL mos_analysis(qs_env_a, qs_env_b)
>
> When I access *qs_env_a* in mos_analysis subroutine I am ending up with 
> CPASSERT failed error in qs_environment_types. 
>
>  *******************************************************************************
>  *   
> ___                                                                       *
>  *  /   
> \                                                                      *
>  * 
> [ABORT]                                                                     
> *
>  *  \___/                             CPASSERT 
> failed                          *
>  *    
> |                                                                        *
>  *  
> O/|                                                                        *
>  * /| 
> |                                                                        *
>  * / \                                              
> qs_environment_types.F:640 *
>
>  *******************************************************************************
>
> My question is how to access the two qs_env and all info associated with 
> it (mos, rho, ks_matrix, etc.) post-scf, in a multiple force_eval 
> calculation before deallocating everything. 
>
> Thank you very much for your help.
>
> Best Regards,
> Pavan.
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20181203/8ffdf862/attachment.htm>