Re: [cclib-devel] Jaguar basis sets

[Noel O'B. <bao...@gm...>]

Thanks for the help. In the end ,I agree and I've just removed Jaguar gbasis.

Rick Muller also checked:
"""
Briefly, I don't have any idea what Jaguar is doing. I just ran H2
using STO-3G with Jaguar, and the basis print out is essentially the
same as that which you reference below. Which looks like it would lead
to 4 bfns for H2, but in fact leads to only 2. Maybe the column
labeled "jconst" isn't what they say it is in the manual. Note that
the basis function index (which I believe is column 6, labeled nfsh)
shows that all three primatives are in the same basis function. But
maybe I'm getting "basis functions" confused with "shells".

I also checked the internal Jaguar basis set data file, and it has
coefficients equivalent to what you quoted from PyQuante (0.154,
0.535, 0.444...). So they just must be printing their output in a
weird way.
"""

Interestingly, Rick points out that the data file *does* have the
correct coefficients.

All gbasis tests now pass.

Noel

On 14/02/07, Karol Langner <kar...@kn...> wrote:
> I tried to look into this yesterday for a bit, but couldn't come up with
> anything else, again, than that there might be an adidtional sum in the
> contraction.
>
> I think eliminating gbasis from the Jaguar parser is the best thing to do for
> the moment, since there is no way be sure what exactly is going on. In the
> long run, it might be possible to get the infromation from the Jaguar people,
> but that not top-priority now, I guess.
>
> Karol
>
> On Tuesday 13 of February 2007 14:29, Noel O'Boyle wrote:
> > You're right - there's still something weird going on, since the total
> > number of basis functions (which is listed in the output file) adds up
> > to the same as in Gaussian. That is, the three GTOs are regarded as a
> > single basis function.
> >
> > I've posted a question to the CCL (no response) and now I'm asking
> > Rick Muller (of PyQuante) in case he has any ideas. I think that if
> > this remains unresolved, we'll just have to exclude Jaguar gbasis and
> > go with the rest.
> >
> > On 12/02/07, Karol Langner <kar...@kn...> wrote:
> > > On Friday 09 of February 2007 09:09, Noel O'Boyle wrote:
> > > > STO-3G is defined differently as a set of 3 GTOs (1S), 2 GTOs + 1 GTO
> > > > (2S), 2GTOs + 1 GTO (2P).
> > > >
> > > > Now that we've figured this out, I will add gbasis to the Jaguar
> > > > parser over the next few days.
> > >
> > > That's a little surprising, that STO-3G is defined differently. Also, the
> > > energy zomes out exactly the same.
> > >
> > > Karol
> > >
> > > --
> > > written by Karol Langner
> > > Mon Feb 12 19:14:36 CET 2007
>
> --
> written by Karol Langner
> Wed Feb 14 10:41:13 CET 2007
>