[Wien] Probelm of the enhanced SOC parallel calculation

[Peter Blaha]

The message makes no sense. The only difference between parallel and 
non-parallel version is that the k-list is splitted and the output files 
have an index like

case.outputso_1, ...

For sure the same executable is used and in the scf file you should 
obtain different results.

On 07/11/2013 10:13 AM, Chang-Jong Kang wrote:
> Dear Prof. Blaha and Wien2k users,
>
> Hello, I would like to enhance the strength of the spin-orbit coupling
> (SOC) in the particular orbital of the particular atom.
>
> For example, two time enhanced SOC in d-orbitals of 7th to 12th atoms is
> wanted.
>
> I slight modified the "hsocalc.f" source file located in the SRC_lapwso
> directory as follows:
>
> ================== hsocalc.f
> =======================================================
>   DO ITY = 1,nat                      !!! line # 139
>             write(*,*) "CJK,ity",ity   !!! CJK modified
>             lfirst=lfirst+mult(ity-1)
>             if (.not.lso(ity)) cycle
>             DO INA = 1,mult(ITY)
>                INDJ=lfirst+ina-1
>                DO L = 1,LABC
>                   DO MF = -L,L
>                      !!----------------------- CJK
> modified------------------------------------------------------
>
> !!-------------------------------------------------------------------------------------------------
>                      if((ITY.ge.7).and.(ITY.le.12)) then
>                        if(L.eq.2) then
>                            write(6,*)
>                            write(6,*) "ITY = ", ITY
>                            write(6,*) "L = ", L
>                            write(6,*) "enhanced SOC"
>                            write(6,*)
>                            lambda_mod=2.0
>                        else
>                            lambda_mod=1.0
>                        endif
>                      endif
>                      !!----------------------- CJK
> modified------------------------------------------------------
>
> !!-------------------------------------------------------------------------------------------------
>                      ILMF = L*L +L +MF +1
>                      DO l_mat2=1,mrf(l,ity)
>                         index=index+1
>
>   hso=hso+dconjg(abcdlm(l_mat2,ilmf,indj,ief,isu))*lambda_mod*&
>   !!! CJK modified
>                              t(index,isf,ibi)
>                      ENDDO
>                   ENDDO
>                ENDDO
>             ENDDO
> ===============================================================================
>
> The problem is,
>
> When I ran the modified one with "run -orb -so", the CJK characters were
> printed and obtained results were reasonable.
> However, when I ran it with "run -orb -so -p", the CJK characters were
> not printed and the results was exactly same as one which was the normal
> strength of the SOC.
>
> This problem might be related to the parallel option.
> I modified also the WIENROOT and PATH in .bashrc, but the situation was
> not improved.
>
> How can I fix the problem?
>
> Any suggestions and comments are appreciated.
>
>
> Best regards,
> Chang-Jong Kang
>
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:  http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
>

-- 

                                       P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300             FAX: +43-1-58801-165982
Email: blaha at theochem.tuwien.ac.at    WWW: 
http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------