[CP2K:1925] Re: problem generating basis set

[Jörg Saßmannshausen]

Hi Matthias,

thanks for your help! I will play with that later :-)

All the best from Glasgow!

Jörg


On Samstag 21 März 2009 Matthias Krack wrote:
> Hi Jörg,
>
> indeed the missing < sign is a typo on my slides. Below you may find
> an example for an input file, which should perform an optimization for
> 6 exponents (family basis set, ie. same exponents for s and p
> orbitals) to facilitate your startup in the Gaussian basis set design
> business. The output file qs.input contains the optimized exponents
> and the corresponding coefficient for the occupied atomic orbitals in
> a convenient format. This output file is not a direct input for CP2K/
> QS, even if the name of the file might imply this. It's on you to
> assemble (and test of course) a proper basis set based on this
> information.
>
> cheers,
>
> Matthias
>
> &ATOM
> NAME  F
> RPRB  2.721
> PPOT
> OPTI  1.0D-8
> ZVAL  9
> ZEFF  7
> LMAX  1
> OCCU
> 1 2
> 1 5
> MIXI  0.2
> CONV  1.0D-14
> ITER  200
> IPPN  100
> XCFN  PADE
> &END
>
> &BASIS
> GAUS
> 6
>        15.4309642052        6.5711763684        2.6219969372
> 1.0693496460
>         0.4217042136        0.1640150222
> 1 2 3 4 5 6
> 6
>        15.4309642052        6.5711763684        2.6219969372
> 1.0693496460
>         0.4217042136        0.1640150222
> 1 2 3 4 5 6
> &END
>
> &POTENTIAL
>     GOEDECKER
> 2                                    LMAX
>       .218524651                                 RC
>   2   -21.307361120     3.072869420   #C  C1 C2
>       .195567211  1  23.584942110 H(s) 11
>       .174268317  0 H(p) no projector
> &END
>
>
> On Mar 21, 3:09 pm, Jörg Saßmannshausen
>
> <jorg.sassm... at strath.ac.uk> wrote:
> > Hi Marius,
> >
> > thanks for the info, now it is working. I simply sticked to the slides
> > (and the redirect sign did not occure to me). Silly me.
> >
> > It is working now, I will search the list for further informations! :-)
> >
> > Have a nice weekend!
> >
> > Jörg
> >
> > On Samstag 21 März 2009 Marius Retegan wrote:
> > > Hi,
> > >
> > > Try adding the redirect sign... like this:
> > > ./optbas < basis-F-q7-g4
> > > Also, those inputs don't optimize neither alpha nor the coefficients.
> > > Search the list, there was someone who explained a little the input.
> > > Basically you should add OPTI 1.0D-7.
> > >
> > > Marius
> > >
> > > On Sat, Mar 21, 2009 at 12:59 AM, Jörg Saßmannshausen
> > > <jorg.sassm... at strath.ac.uk<mailto:jorg.sassm... at strath.a
> > >c.uk
> > >
> > >>> wrote:
> > >
> > > Dear all,
> > >
> > > I have a problem generating the basis set for F and the PBE functional.
> > > Following Matthias' talk at the Tutorial, I did (in the cp2k
> > > directory):
> > >
> > > - cvs co basis_sets
> > > - cd basis_sets/source
> > >
> > > I then had a look at the Makefile, change the path to blas etc. I was
> > > using the g95-32 compiler to compile optbas. The compilation worked
> > > without any problems.
> > >
> > > I then was looking in the inpout folder for a suitable script for F and
> > > I found the basis-F-q7-g4 one.
> > > Just to make sure I am not messing up things, I copied basis-F-q7-g4
> > > and optbas in a different folder. Trying to follow the tutorial, I did
> > >
> > > ./optbas basis-F-q7-g4
> > >
> > > which basically stopped here:
> > >
> > > PROGRAM ATOM
> > > reading input file...
> > >
> > > end of story, nothing happend. So I decided to try the Intel Fortran
> > > Compiler, recompiled, and now I get:
> > >
> > >       PROGRAM ATOM
> > >        ------------
> > >  reading input file...
> > > forrtl: Illegal seek
> > > forrtl: severe (20): REWIND error, unit 5, file /dev/pts/2
> > > Image              PC                Routine            Line       
> > > Source optbas             000000000057B832  Unknown              
> > > Unknown  Unknown optbas             000000000057A7C4  Unknown          
> > >     Unknown  Unknown optbas             000000000052F6BA  Unknown      
> > >         Unknown  Unknown optbas             00000000004FC743  Unknown  
> > >             Unknown  Unknown optbas             00000000004FC030 
> > > Unknown               Unknown  Unknown optbas            
> > > 000000000050FE96  Unknown               Unknown  Unknown optbas        
> > >     000000000041667D  Unknown               Unknown  Unknown optbas    
> > >         0000000000427B9B  Unknown               Unknown  Unknown optbas
> > >             000000000040D2FC  Unknown               Unknown  Unknown
> > > optbas             00000000004062A2  Unknown               Unknown 
> > > Unknown libc.so.6          00002ADF44335154  Unknown              
> > > Unknown  Unknown optbas             00000000004061E9  Unknown          
> > >     Unknown  Unknown
> > >
> > > Changing the optimisation level from -O3 to -O2 did not help. The
> > > compiler version is 10.1 20081024, the same compiler which successfully
> > > builds cp2k.
> > >
> > > Ok, and this is where I am stuck. Is there a human input error or is
> > > there some other problem. I tried google, but other than Matthias' talk
> > > did not produced something usefull.
> > >
> > > Could somebody just give me a quick help and point me in the right
> > > direction?
> > >
> > > Many thanks.
> > >
> > > All the best from Glasgow
> > >
> > > Jörg
> > >
> > > --
> > > *************************************************************
> > > Jörg Saßmannshausen
> > > Research Fellow
> > > University of Strathclyde
> > > Department of Pure and Applied Chemistry
> > > 295 Cathedral St.
> > > Glasgow
> > > G1 1XL
> > >
> > > email:
> > > jorg.sassm... at strath.ac.uk<mailto:jorg.sassm... at strath.ac
> > >.uk> web:http://sassy.formativ.net
> > >
> > > Please avoid sending me Word or PowerPoint attachments.
> > > Seehttp://www.gnu.org/philosophy/no-word-attachments.html
> >
> > --
> > *************************************************************
> > Jörg Saßmannshausen
> > Research Fellow
> > University of Strathclyde
> > Department of Pure and Applied Chemistry
> > 295 Cathedral St.
> > Glasgow
> > G1 1XL
> >
> > email: jorg.sassm... at strath.ac.uk
> > web:http://sassy.formativ.net
> >
> > Please avoid sending me Word or PowerPoint attachments.
> > Seehttp://www.gnu.org/philosophy/no-word-attachments.html
>
> 


-- 
*************************************************************
Jörg Saßmannshausen
Research Fellow
University of Strathclyde
Department of Pure and Applied Chemistry
295 Cathedral St.
Glasgow
G1 1XL

email: jorg.sassm... at strath.ac.uk
web: http://sassy.formativ.net

Please avoid sending me Word or PowerPoint attachments.
See http://www.gnu.org/philosophy/no-word-attachments.html