Firefly and PC GAMESS-related discussion club


 
Learn how to ask questions correctly  
 
 
We are NATO-free zone
 



Re^3: SA2-CAS(2,2) calculation with cistep=ALDET and ISPIN=0

Thomas Pijper
thomaspijper@hotmail.com


Dear Evgeniy,

NSTATE merely selects the number of CI states to take into consideration. It does not specify which states to average over. Instead, the WSTATE keyword can be used for this.

As an example, if the first three states are of the multiplicities [singlet, triplet, singlet] and you would like to average over the two singlet states, specify the following:

NSTATE=3 WSTATE(1)=1,0,1

Hope this helps.


Kind regards,
Thom

On Mon Feb 27 '17 9:38pm, GrEv wrote
------------------------------------
>Thanks for your response. Yes I set NSTATE to 2, because I want to do SA2-CASSCF(2,2) averaging over two singlet states S0 and S1. But this seems to be not possible with cistep=guga, because ISPIN doesn't work here, and without ISPIN one cannot select two singlet states, as the second state (after S0) is a triple.

>Best regards,
>Evgeniy

>On Sat Feb 25 '17 10:29pm, Alex Granovsky wrote
>-----------------------------------------------
>>Hello,

>>the ISPIN option works only with iterative Davidson eigensolver.

>>With CASSCF(2,2) and alike, Hamiltonian matrix is tiny and is
>>diagonalized directly using one of available diagonalization
>>routines rather than iteratively, so that the ISPIN option is
>>ignored.

>>My guess is that you set NSTATE to 2 and got two states,
>>namely lowest singlet and lowest triplet. This is why program
>>complains (as it needs two singlet states). Use NSTATE=3 and
>>you'll get two singlets and a triplet state.

>>Kind regards,
>>Alex Granovsky
>>
>>
>>
>>
>>On Fri Feb 24 '17 2:18pm, GrEv wrote
>>------------------------------------
>>>Hello,

>>>I have a question related to a SA2-CAS(2,2) calculation using cistep=ALDET and ISPIN=0.
>>>I am interested in two singlet states, S0 and S1. Even with NSTGSS=4
>>>the calculation crashes with the error:

>>> ***** ERROR *****
>>> THIS RUN FOUND    1 CI EIGENVECTORS WITH S= 0.00,
>>> BUT YOU REQUESTED STATE AVERAGING OF    2 ROOTS.
>>> PLEASE EXAMINE YOUR CHOICE OF -NSTATE- INPUT DATA.

>>>So, I wonder if cistep=ALDET can be used in such a case, or only cistep=GUGA works here?

>>>Best regards,
>>>Evgeniy


[ Previous ] [ Next ] [ Index ]           Wed Mar 1 '17 0:11am
[ Reply ] [ Edit ] [ Delete ]           This message read 184 times