[CPMD-list] velocity output

[Ari P Seitsonen]

Dear Xiandong, Axel et al,

   Ehmm, actually this has been implemented but not documented... Please 
check the options 'SMALL' and 'RANGE' for 'TRAJECTORY'. Even more 
"dangerous" options include 'BIN'... But like Axel says, it is better to 
think carefully before starting to reduce the amount of data, as one might 
regret afterwards the need of some information that was left out.

   The ultimate reference: Source code files 'control.F'

--------------------------------------------------------
  ...
         ELSE IF(INDEX(LINE,'TRAJ').NE.0) THEN
C..Store trajectory
           ROUT=.TRUE.
           IF(INDEX(LINE,'XYZ').NE.0) XTOUT=.TRUE.
           IF(INDEX(LINE,'DCD').NE.0) DCOUT=.TRUE.
           IF(INDEX(LINE,'OFF').NE.0) THEN
             ROUT=.FALSE.
             XTOUT=.FALSE.
             DCOUT=.FALSE.
           ENDIF
           IF(INDEX(LINE,'BIN').NE.0) THEN
             TWRITEBINTRAJECTORY=.TRUE.
           ENDIF
           IF(INDEX(LINE,'FORCE').NE.0) THEN
             TWRITEFORCETRAJECTORY=.TRUE.
           ENDIF
           IF(INDEX(LINE,'SMALL').NE.0) TRAJSMALL=.TRUE.
           IF(INDEX(LINE,'RANGE').NE.0) THEN
             LINEOLD=LINE
             READ(IUNIT,ERR=99,END=99,FMT=*)
      $           MINWRITEATOM, MAXWRITEATOM
           END IF
  ...
-------------------------------------------------------

and 'printp.F'.

     Greetings,

        apsi

PS The "range of atoms" was written when playing with the EAM code in
   CPMD: The amount of atoms was maybe thousands, the number of iterations
   millions and we were studying a surface where the bottom of the slab was
   fixed => the coordinates would have been identical at each step, their
   velocities being exactly zero

-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
   Ari P Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/
   IMPMC, CNRS & Université Pierre et Marie Curie
   Tel: +33-1-4427 7542, Fax: +33-1-4427 3785, GSM: +33-6-6736 3820

On Fri, 7 Sep 2007, Axel Kohlmeyer wrote:

> On Fri, 7 Sep 2007, liu liuxiandong wrote:
>
>
> XL> Dear CPMDers,
>
> dear xiandong liu,
>
> XL>   I want to extract the vibration properties of some atom species from the
> XL> velocity correlation function along the MD simulation. I find that the
> XL> output TRAJECTORY file contains all ionic positions and velocities, so the
> XL> file is huge for a MD trajectory lasting several picoseconds. Could CPMD
> XL> outputs the velocities of the atom type specified by the user ? Any
> XL> suggestion is appreciated .
>
> in principle this is possible, but you have to modify the sourcecode
> for that. however, is wonder whether this is worth the effort.
> if you can afford to run a long trajectory of a large system, so that
> the total storage requirement would really matter, you can also afford
> to have extra (external?) disk storage for keeping and analysing your
> trajectory. in comparison to equivalently costly classical MD
> calculations or even coarse-grained MD, the disk size requirements for
> a CPMD trajectory will always be miniscule.
>
> please keep in mind, that it is good practice to split CPMD runs into
> segments and then build a continous trajectory through restarting.
> if you rename the outputs/files after each segment (which is actually
> good practice and highly recommended) you will have a series of
> shorter files that will be easier to post-process.
>
> cheers,
>   axel.
>
>
> XL>
> XL>
> XL>
> XL> Xiandong Liu
> XL>
>
> -- 
> =======================================================================
> Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
>   Center for Molecular Modeling   --   University of Pennsylvania
> Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
> tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
> =======================================================================
> If you make something idiot-proof, the universe creates a better idiot.
>
> _______________________________________________
> CPMD-list mailing list
> CPMD-list at cpmd.org
> http://cpmd.org/mailman/listinfo/cpmd-list
>