UGBS2P

Axel akoh... at gmail.com
Thu Aug 28 04:42:49 UTC 2008



On Aug 27, 10:24 pm, Ahmad <ahmad.... at gmail.com> wrote:
> Hello everybody,

hello ahmand,

> I want to examine the basis set dependence on my calculations. So I
> need to carry out the calculations using the very extended UGBS2P
> Gaussian type basis set.But I could not find this basis set file from
> cp2k/tests/QS/ route.

please repeat after me: cp2k is not gaussian.

if you are using quickstep you need basis sets
that are adapted to the pseudopotentials and
functionals that you are using. what about
incrementally stepping up the existing basis sets
and looking for convergence?

please note, that the density cutoff is equally important
for accurate results. but also keep in mind that oversized
basis sets are a waste of computer time since you are limited
by the systematic errors of the method, i.e. DFT in this case.

> Could you please tell me how to find the basis set file. Do you have
> this file?

the header of the basis set file describes the format.
feel free to enter the basis set for yourself.

cheers,
   axel.

> Your helps are greatly appreciated.
>
> Sincerely yours,
>
> Ahmad


More information about the CP2K-user mailing list