[Columbus] argos problem -- done

[Scott Brozell]

Hi Werner,

Thanks for following up.
Although this might be considered a feature, I shall submit a bug report.
Would you send your before and after argosin files ?

thanks,
Scott Brozell

On Thu, 1 Sep 2005, Werner Gyorffy wrote:

> Hi all,
>
> Some times ago I posted a question about an argos problem. I would like to
> say thanks to all who replied, and report the solution I found.
>
> I wrote:
>
> > Hi all,
> >
> > I would like to perform a calculation (CASSCF) for acetylene using an ANO
> > basis /Widmark et al TCA 79(1991)419/ but I have technical
> > problem with the
> > input. 3s2p1d contraction is used for C, and 3s for the H atoms. iargos
> > finished everything fine but argos did not work. First, I had to
> > change the
> > parameter mconup to 28 in argos2.f, suggested by the program. I recompiled
> > the standard package with -nox and I tried to run argos again. I got a
> > message as follows.
> >
> > PGFIO-F-225/list-directed read/unit=5/lexical error-- unknown token type.
> > File name = argosin    formatted, sequential access   record = 118
> > In source file argos2.f, at line number 1033
> >
> > Could anyone please help me. (Maybe it is something trivial...)
> >
> > Thanks,
> >
> > Werner
>
> I wanted to use a several contractions from the "Widmark et al TCA
> 79(1991)419" ANO basis, namely, [3s,2p,1d] out of the [7s,7p,4d,3f] for
> carbon. In the argos input generation I selected 3 of the 7 s, 2 of the 7 p,
> 1 of the 4 d, and 0 (!) of the 3 f functions. I thought that 0 for the
> number of f functions would get rid of the f contraction set totally. It
> turned out that even if I specified 0 f functions I needed to DELETE the f
> contraction set afterwards. Now everything is working fine.
>
> Sorry for that if it was an obvious thing to you.