On 2010-07-08, at 11:04 AM, Andrey Frolov wrote:
> Dear NAB and 3D-RISM Developers,
>
> I have calculated the system of TIP3P water, Cs and Cl ions. So i have 4
> atoms: O, H, Cs and Cl. They can be indexed by: S1A1, S1A2, S2A1, S3A1.
> (where S stands for SPECIES and A for ATOMS).
>
> It seems to me that the atom ordering in the potential output is
> incorrect. When in the input to 1drism i put " OUTLST='xU' ", i get in
> the output.uvv file the order of atom pairs potential:
> ivv iat1 iat2
> 1 S1A1 S1A1
> 2 S1A1 S1A2
> 3 S1A1 S2A1
> 4 S1A1 S3A1
> 5 S1A2 S1A2
> 6 S1A2 S2A1
> 7 S1A2 S3A1
> 8 S2A1 S2A1
> 9 S2A1 S3A1
> 10 S3A1 S3A1
>
> This order in output was generated with the following loops:
> ivv=0
> do isp=1,nsp
> do iat=1,nat(isp)
> do isp2=isp,nsp
> do iat2=1,nat(isp2)
> ...
> enddo
> enddo
> enddo
> enddo
>
This ordering is only used for the separation heading in Uvv output file and nowhere else. It is incorrect.
>
> BUT!!! The true potential order is generated with these loops:
> ivv = 0
> do iv2=1,nv
> do iv1=1,iv2
> ivv = ivv + 1
> ....
> enddo
> enddo
>
> Which gives the following TRUE ordering:
> ivv iat1 iat2
> 1 S1A1 S1A1
> 2 S1A2 S1A1
> 3 S1A2 S1A2
> 4 S2A1 S1A1
> 5 S2A1 S1A2
> 6 S2A1 S2A1
> 7 S3A1 S1A1
> 8 S3A1 S1A2
> 9 S3A1 S2A1
> 10 S3A1 S3A1
>
>
> I very hope that this is mixed up only in the potential output.
>
This is the true order of the output for Uvv and Gvv and is the order used internally for all calculations.
> Could it affect the order of susceptibility functions in the input for
> 3D RISM?
No, this has nothing to do with the susceptibility, which is probably why the incorrect labelling was missed. The susceptibility is output with the full interaction matrix and not the triangular form used for Uvv and Gvv.
Tyler
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Jul 09 2010 - 08:30:03 PDT
