[CP2K-user] [CP2K:14481] Regular grid of gaussians

[hut... at chem.uzh.ch]

Hi

you will need to define an AUX_BASIS also for the ghost element.
Your on your own to find a reasonable basis that is compatible
with your ghost basis. 

Just a wild guess. If you are using a grid of basis functions, your
basis will be very small (single Gaussian, maybe s and p types).
In that case use the same basis for the AUX_BASIS.

Other idea: try to use no ADMM basis on the grid.
            Method 1) try AUX_BASIS NONE
            Method 2) Use ADMM METHOD BLOCKED_PROJECTION

I have never tried any of these, you need to experiment.

regards

Juerg Hutter

--------------------------------------------------------------
Juerg Hutter                         Phone : ++41 44 635 4491
Institut für Chemie C                FAX   : ++41 44 635 6838
Universität Zürich                   E-mail: hut... at chem.uzh.ch
Winterthurerstrasse 190
CH-8057 Zürich, Switzerland
---------------------------------------------------------------

-----cp... at googlegroups.com wrote: -----
To: "cp2k" <cp... at googlegroups.com>
From: "Paolo Pegolo" 
Sent by: cp... at googlegroups.com
Date: 01/05/2021 04:25PM
Subject: Re: [CP2K:14481] Regular grid of gaussians

Ok, thank you. I see that the required ELEMENT keyword changes only the atomic label, in this case.
If I want to use ADMM, will I have to use the same basis set also for the AUX_BASIS part? Or is there a way to generate a consistent auxiliary basis?

Regards, 

Paolo

Il giorno martedì 5 gennaio 2021 alle 10:44:22 UTC+1 jgh ha scritto:
Hi 
 
You can specify an element type using the keyword ELEMENT XX 
in the KIND section.  
 
regards 
 
Juerg Hutter 
-------------------------------------------------------------- 
Juerg Hutter                         Phone : ++41 44 635 4491 
Institut für Chemie C                FAX   : ++41 44 635 6838 
Universität Zürich                   E-mail: h... at chem.uzh.ch 
Winterthurerstrasse 190 
CH-8057 Zürich, Switzerland 
--------------------------------------------------------------- 
 
-----c... at googlegroups.com wrote: ----- 
To: "cp2k" <c... at googlegroups.com> 
From: "Paolo Pegolo"  
Sent by: c... at googlegroups.com 
Date: 01/04/2021 02:43PM 
Subject: Re: [CP2K:14469] Regular grid of gaussians 
 
Thank you for your answer. 
 
I wrote a file MY-BASIS with the new basis for the element AGHOST with a basis set called my-basis-set. The beginning of such file reads 
  AGHOST my-basis-set 
 
Then I added 
 
BASIS_SET_FILE_NAME MY-BASIS 
 
to the &DFT section of the input file,  
 
&KIND AGHOST 
BASIS_SET my-basis-set 
GHOST 
&END KIND 
 
to the &SUBSYS section, and I put the coordinates of the AGHOST atoms I need to the XYZ file I use to define the topology of the system. I get the error  
 
Element <AGHOST> provided for KIND <AGHOST> which cannot be mapped 
 with any standard element label. Please correct your input file!  
 
Is there something wrong I am doing? 
 
Regards, 
Paolo  
 
Il giorno lunedì 4 gennaio 2021 alle 10:11:52 UTC+1 jgh ha scritto: 
Hi  
  
You should be able to do such a calculation by defining a new   
atom type, e.g. AGHOST  
  
    &KIND AGHOST  
      BASIS_SET "my-basis-set"  
      GHOST  
    &END KIND  
  
Then add atoms of this type on a regular grid to the coordinate definition.  
  
For a test I would start with a small number of ghost atoms and check  
the number of basis functions and the outcome of a geometry optimization  
for a simple molecule.  
  
regards  
  
Juerg Hutter  
  
--------------------------------------------------------------  
Juerg Hutter                         Phone : ++41 44 635 4491  
Institut für Chemie C                FAX   : ++41 44 635 6838  
Universität Zürich                   E-mail: h... at chem.uzh.ch  
Winterthurerstrasse 190  
CH-8057 Zürich, Switzerland  
---------------------------------------------------------------  
  
-----c... at googlegroups.com wrote: -----  
To: "cp2k" <c... at googlegroups.com>  
From: "Paolo Pegolo"   
Sent by: c... at googlegroups.com  
Date: 01/03/2021 02:50PM  
Subject: [CP2K:14462] Regular grid of gaussians  
  
Dear cp2k community,  
  
In the literature (e.g., here and here) people refer to the possibility of adding to the set of atomic centered gaussians a grid of gaussians functions fixed in space, that helps describing the wavefunction of a system with, e.g., loosely bound electrons. They use cp2k, but I have not been able to find any information about it in the manual.  
  
How can one do this?  
  
Thank you,   
Paolo    
  --   
 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 cp... at googlegroups.com.  
 To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/09b518d1-6cbf-4b83-ae5e-53a2bee76125n%40googlegroups.com.  
   
   
  --  
 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 cp... at googlegroups.com. 
 To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/e52e835d-8276-4355-8320-c2a8129b03ean%40googlegroups.com. 
  
  
  -- 
 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 cp... at googlegroups.com.
 To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/4b1f4318-adf3-4ea3-99c7-bf08c7063086n%40googlegroups.com.