[CP2K-user] [CP2K:21939] Re: density fitting auxbasis in cp2k

[Frederick Stein]

Hi,
I am not quite sure which MOLOPT auxiliary basis sets you are referring to. 
CP2K does not have preoptimized RI basis sets for HFX but for RI-MP2, 
RI-RPA, GW and separate ones for the LRIGPW-approach only. The 
recommendation is to use automatically generated basis instead 
(see https://manual.cp2k.org/cp2k-2025_2-branch/methods/dft/hartree-fock/ri_gamma.html). 
In case of all-electron calculations, you can employ those corresponding to 
the primary basis set from https://www.basissetexchange.org/.
HTH,
Frederick
yj jiang schrieb am Donnerstag, 30. Oktober 2025 um 14:21:33 UTC+1:

> Hello everyone,
>
> I am implementing Density Fitting (DF) for Hybrid DFT using the MOLOPT 
> basis set.
>
> My reference, PySCF, uses the Even-Tempered Basis (ETB) to generate the 
> auxiliary basis set (auxbasis). This ETB is extremely large, often over 10 
> times the size of the primary basis, making the calculation highly 
> inefficient.
>
> I am seeking a method to generate a more compact auxbasis. I know MOLOPT 
> basis come from CP2K, so come here for some help.
>
> Could anyone provide references or specific details on:
>
> CP2K's strategy for generating these compact MOLOPT auxiliary basis sets 
> (i.e., the recipe that avoids the oversized ETB)?
>
> Any specific parameters (like geometric progression rules) that could 
> replace the default PySCF ETB for MOLOPT while maintaining accuracy?
>
> Thanks for any guidance!

-- 
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+unsubscribe at googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/cp2k/7c8fb71e-e595-4527-83c0-f4a983bff232n%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20251030/f604e70b/attachment.htm>