Groups keyboard shortcuts have been updated
Dismiss
See shortcuts

odd behavior from X-AMFI

58 views
Skip to first unread message

Peterson, Kirk A

unread,
Apr 3, 2025, 11:48:41 AMApr 3
to dirac-users, Bonar, Brooke

Dear dirac experts (and probably more specifically Trond Saue in this case):

 

When using the new X-AMFI routines on the U atom (in preparation for a molecular calculation), the 4c DHF part of the X2Cmmf calculation converges correctly, but after the 2c transformation the character of the open shell mj=1/2 spinor in gerade symmetry is displayed incorrectly in the Mulliken analysis as a U 7s spinor rather than a 6d.  The spinor energy is correct for the 6d. The output that is attached to this email was started from a CHECKPOINT.h5 file corresponding to a 4c DHF calculation, so the X2Cmmf calculations converges in 2 iterations.

 

Best regards,

 

-Kirk

dhf_amf_vtz.out

Trond SAUE

unread,
Apr 3, 2025, 11:53:21 AMApr 3
to dirac...@googlegroups.com

Hi Kirk,

I can look at this, but it will only be early next week.

All the best

    Trond

--
Trond Saue
Laboratoire de Chimie et Physique Quantiques
UMR 5626 CNRS --- Université Toulouse III-Paul Sabatier
118 route de Narbonne, F-31062 Toulouse, France

Phone : +33/561556361 Fax: +33/561556065
Mail : trond...@irsamc.ups-tlse.fr
Web : https://dirac.ups-tlse.fr/saue
DIRAC : http://www.diracprogram.org/
ESQC : http://www.esqc.org/
ERC Advanced Grant: HAMP-vQED
Book: Principles and Practices of Molecular Properties: Theory, Modeling, and Simulations

peterso...@gmail.com

unread,
Apr 3, 2025, 11:57:41 AMApr 3
to dirac...@googlegroups.com

Hi Trond,

 

Thank you, that would be great.  Did you see my previous email to the list about .eamf and .gaunt ?

 

Best regards,

 

-Kirk

 

--
You received this message because you are subscribed to the Google Groups "dirac-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dirac-users...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/dirac-users/3d328dd6-68de-4249-b719-4782df7ed731%40irsamc.ups-tlse.fr.

Peterson, Kirk A

unread,
Apr 3, 2025, 1:31:00 PMApr 3
to dirac...@googlegroups.com, Bonar, Brooke

As an addendum to what I note below, when one does a X-AMFI calculation with the amf option in the molecule, the convergence of the DHF is very much spoiled. I can give an initial guess from a converged X2C calculation (old version) and after 100 iterations it (sort of ) converges to something very different (and not correct).  EAMF doesn’t show this issue.

 

Regards,  -Kirk

 

--

You received this message because you are subscribed to the Google Groups "dirac-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to

ccsd_amf_vtz.out
ccsd_eamf_vtz.out

Peterson, Kirk A

unread,
Apr 3, 2025, 2:31:43 PMApr 3
to dirac...@googlegroups.com

Dear Trond,

 

Sorry for the bunch of questions today on X-AMFI (obviously I’m really interested in using it!).  One more for you:  what is the appropriate way to carry out an atomic, open-shell CCSD calculation with X-AMFI and eamf (or amf) that is consistent with a molecular calculation, e.g., for calculating a bond dissociation energy?  If I use the same approach as a molecule:

 

**HAMILTONIAN

.X2C

.DOSSSS

*X-AMFI

.eamf

 

Using a previously determined amfPCEC.h5 file, the calculation dies with:

 

ERROR: trying to read non-existent data set HF/9/custom/finite//aooDM_4c from amfPCEC.h5

 

If instead I use (same as the generation of the amfPCEC.h5 files):

 

**HAMILTONIAN

.X2Cmmf

.DOSSSS

*X-AMFI

.eamf

 

the CCSD does not converge since it doesn’t recompute the Fock matrix in an X2Cmmf calculation (usual problem I have for open-shell calculations with X2Cmmf).

 

Best regards,  -Kirk

 


Date: Thursday, April 3, 2025 at 8:53 AM
To: dirac...@googlegroups.com <dirac...@googlegroups.com>

--

You received this message because you are subscribed to the Google Groups "dirac-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to

Trond SAUE

unread,
Apr 3, 2025, 5:18:03 PMApr 3
to dirac...@googlegroups.com

Dear Kirk,

I note your questions and will try to answer ASAP next week.

All the best,

peterso...@gmail.com

unread,
Apr 3, 2025, 5:23:26 PMApr 3
to dirac...@googlegroups.com

Dear Trond,

 

Thanks in advance!

 

Best wishes,

 

-Kirk

 

From: dirac...@googlegroups.com <dirac...@googlegroups.com> on behalf of Trond SAUE <trond...@irsamc.ups-tlse.fr>
Date: Thursday, April 3, 2025 at 2:18 PM
To: dirac...@googlegroups.com <dirac...@googlegroups.com>
Subject: Re: [dirac-users] odd behavior from X-AMFI

--

You received this message because you are subscribed to the Google Groups "dirac-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dirac-users...@googlegroups.com.

Trond SAUE

unread,
Apr 9, 2025, 5:44:59 AMApr 9
to dirac...@googlegroups.com

Hi Kirk,

I am looking into your case. In the input I find

.CLOSED SHELL
 44 44
.OPEN SHELL
 2
 1/4,0
 3/0,6
#s1/2 p1/2 p3/2 d3/2 d5/2 f5/2 f7/2 g7/2 g9/2
.KPSELEC
7
 -1   1   -2    2   -3    3   -4
 14  10   20   12   18    6    8
  0   0    0    4    0    0    0
  0   0    0    0    0    6    0

Uranium has ground state configuration [Rn]5f^3_6d^1_7s^2.

The above specification is [Rn]5f_{5/2}^3 6d_{3/2}^1 7s_{1/2}^2,

so you can see it is sort of the same configuration, but only allowing electrons in 5f_{5/2} and not 5f_{7/2}, likewise only allowing electrons in 6d_{3/2} and not 6d_{5/2}.

Any particular reason for this ?

Trond SAUE

unread,
Apr 9, 2025, 5:51:22 AMApr 9
to dirac...@googlegroups.com

Hi again,

do you have the basis set file cc-pVTZ-DK3 ?

peterso...@gmail.com

unread,
Apr 9, 2025, 9:39:59 AMApr 9
to dirac...@googlegroups.com

Hi Trond,

 

Yes, if one restricts the occupation to just the lower J stacks of the 6d and 5f, the ground state of U is very single determinantal and coupled cluster works very well.

 

Best,  -Kirk

 

From: dirac...@googlegroups.com <dirac...@googlegroups.com> on behalf of Trond SAUE <trond...@irsamc.ups-tlse.fr>
Date: Wednesday, April 9, 2025 at 2:44 AM
To: dirac...@googlegroups.com <dirac...@googlegroups.com>
Subject: Re: [dirac-users] odd behavior from X-AMFI

--

You received this message because you are subscribed to the Google Groups "dirac-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dirac-users...@googlegroups.com.

peterso...@gmail.com

unread,
Apr 9, 2025, 9:41:31 AMApr 9
to dirac...@googlegroups.com

Hi Trond,

 

See attached (although very similar results are obtained with Ken’s TZ basis set).

 

Best,  -Kirk

 

 

From: dirac...@googlegroups.com <dirac...@googlegroups.com> on behalf of Trond SAUE <trond...@irsamc.ups-tlse.fr>
Date: Wednesday, April 9, 2025 at 2:51 AM
To: dirac...@googlegroups.com <dirac...@googlegroups.com>
Subject: Re: [dirac-users] odd behavior from X-AMFI

--

You received this message because you are subscribed to the Google Groups "dirac-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dirac-users...@googlegroups.com.

cc-pVTZ-DK3.txt

Trond SAUE

unread,
Apr 10, 2025, 11:42:44 AMApr 10
to dirac...@googlegroups.com

Dear Kirk,

just letting you know that I work on your case: https://gitlab.com/dirac/dirac/-/issues/131

peterso...@gmail.com

unread,
Apr 10, 2025, 1:44:48 PMApr 10
to dirac...@googlegroups.com

Dear Trond,

 

Thank you very much, I appreciate it.

 

Best,  -Kirk

 

From: dirac...@googlegroups.com <dirac...@googlegroups.com> on behalf of Trond SAUE <trond...@irsamc.ups-tlse.fr>
Date: Thursday, April 10, 2025 at 8:42 AM
To: dirac...@googlegroups.com <dirac...@googlegroups.com>
Subject: Re: [dirac-users] odd behavior from X-AMFI

--

You received this message because you are subscribed to the Google Groups "dirac-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dirac-users...@googlegroups.com.

Peterson, Kirk A

unread,
May 1, 2025, 7:49:39 PMMay 1
to dirac...@googlegroups.com

Dear Trond,

 

Were you also able to reproduce our issues with EAMF and Gaunt, i.e., inclusion of the latter results in:

 

Gaunt not implemented for SOFOCK

 

This happens before the 1st iteration of the DHF. 

 

Best regards,  -Kirk

 

 

From: dirac...@googlegroups.com <dirac...@googlegroups.com> on behalf of Trond SAUE <trond...@irsamc.ups-tlse.fr>
Date: Thursday, April 10, 2025 at 8:42 AM
To: dirac...@googlegroups.com <dirac...@googlegroups.com>
Subject: Re: [dirac-users] odd behavior from X-AMFI

--

Trond Saue

unread,
May 1, 2025, 8:54:44 PMMay 1
to dirac...@googlegroups.com
Dear Kirk, 
I am on your case, but right now on vacation in Guadeloupe.
All the best,
Trond

Sent from my iPad

On 1 May 2025, at 19:49, 'Peterson, Kirk A' via dirac-users <dirac...@googlegroups.com> wrote:



peterso...@gmail.com

unread,
May 1, 2025, 9:08:02 PMMay 1
to dirac...@googlegroups.com

Dear Trond,

 

Sorry to disturb you on vacation!  Have a great time.

 

Best regards,

 

-Kirk

 

Reply all
Reply to author
Forward
0 new messages