Ben Levine ben.l... at gmail.com
Fri Dec 5 16:23:39 UTC 2008

Hi Teo,
Thanks for the VERY quick reply!  I'll update and get back to work.  I
had been running jobs with Zn for several weeks now with an older
version of cp2k, but given your explanation I'm not sure how.


On Dec 5, 11:20 am, Teodoro Laino <teodor... at gmail.com> wrote:
> Hi Ben,
> thanks for reporting this bug. It's fixed now in the CVS.
> The reason is that internally elements internally are stored uppercase
> and when reading parameters with DFTB
> there was no uppercase conversion for the element read from the DFTB
> parameter file.
> I'm just a bit suspicious whether this input may have never run with
> CP2K. The uppercase call was
> missing since the very beginning and no change has been done in the last
> 2-3 years on the convention to store
> internally element as uppercase.
> Anyway.. now your input is working.
> Thanks for reporting.
> Ciao,
> Teo
> Ben Levine wrote:
> > Hi Guys,
> > I recently updated and recompiled my copy of CP2K, and I found a
> > difficulty I previously didn't have.  I've been running some SCC DFTB
> > jobs using the parameters I took from the "tests" subdirectory.  Since
> > my recent update my jobs all fail with the messages:
> > ERROR3 in qs_dftb_parameters:qs_dftb_param_init processor
> > condition FAILED at line 198
> > I dug around in the code a little and found that this is caused
> > because in one place the element Zinc is stored as "Zn" and in another
> > it is stored as "ZN", and these don't match up.  I tried to change the
> > name in my xyz input file, but this didn't change the problem.  I'll
> > upload a sample job which reproduces this problem in the following
> > files:
> > Zn6b.inp
> > Zn6b.xyz
> > Thanks for your help!
> > Ben

More information about the CP2K-user mailing list