[CPMD-list] problem with path integrals restart

[Axel Kohlmeyer]

hi again,

just to finish this up.

i (re-)ran the the tests on my machine and they work for me with
either version 3.11 or the current cvs code. so it is probable that
this is caused by a miscompiled executable and/or library.

On Wed, Apr 2, 2008 at 9:39 AM, Serge Ivanov
<sergei.ivanov at theochem.ruhr-uni-bochum.de> wrote:
>   Dear Marek,
>
>  Few things concerning your inputs. Always try to remove keywords, that
>  remain from other setups (like DEBROGLIE CENTROID in your optimization
>  input), even if you think that they should be ignored. We already have
>  seen cases where such "doing nothing" keywords led to very nasty and
>  hard to figure problems.
>
>  Second thing is the QUENCH BO that you are using - try without it. You
>  already restart from wavefunction optimization results and thus you don't
>  need it. I remember problems with QUENCH BO, some of them were solved, but

actually, in QUENCH BO for path integral is not implemented at all in the
released version (or the next). there should be a note in the output,
right after
the pseudopotential listing.

>  I suspect that those patches will appear only in the coming CPMD release.

??????,
    axel.


>  Last thing that I wanted to mention: did you ensure that your LATEST file
>  contains the proper RESTART_P.1 (where P is the number of beads) ?
>
>  WBR, Serge

-- 
=======================================================================
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.