Firefly and PC GAMESS-related discussion club

 
Learn how to ask questions correctly  
 
 

 

Alex Granovsky
gran@classic.chem.msu.su

On Thu Jun 21 '12 11:22pm, Solntsev Pasha wrote
-----------------------------------------------
>Dear Alex.

>Thank you for your explanation. Now it is clear. I just wont to clarify one more thing. In an outout file for nonabelian groups FF prints something like this (group=C4v istsym=1)

> THE ELECTRONIC STATE IS  2-A1  
> THE ELECTRONIC STATE IS  2-B1  

> UNABLE TO MAKE AN UNAMBIGUOUS DETERMINATION OF THE
> SPATIAL SYMMETRY IN THIS NONABELIAN GROUP.

actually, GUGA code is limited to abelian symmetry groups only.
While it is still possible to specify some non-abelian groups,
they will be mapped to the largest abelian subgroup. In this case,
some irreps of non-abelian group may correspond to several
different irreps of abelian group used.  The code lists all these
irreps. If there is more than one irrep listed, it prints this
warning message.


>Does it mean that states of A1 and B1 symmetry will be only taken into account?

Yes, it does.

>And "UNAMBIGUOUS DETERMINATION" is because we have transition between E->E orbitals. So, program can't correctly determine symmetry of the formed excited state. Am i right?

No, the things are actually much simpler, see above.

>Is it ok to do SA-CASSCF over states of the symmetry say A1(C4v) and E(C4v) in two different runs to get information about excited states. I don't think this is good idea, but may be i am wrong.

It would be better to use a common set of orbitals. After performing
some preliminary computations, you may disable the use of symmetry
inside GUGA code by setting group to c1, and then state-average over
only the desired set of states. You may want to turn state tracking
on as well.

All the best,
Alex

>Thank you for your time.

>Pavel.
>
>
>
>On Wed Jun 20 '12 11:33pm, Alex Granovsky wrote
>-----------------------------------------------
>>Dear Pavel,

>>Sorry, negative values of istsym in GUGA code is just a bug in the program
>>which does not check for non-positive istsym properly.
>>I've just fixed this. Allowed values for istsym are 1..irpmax where
>>irpmax is the total number of irreducible representations of the
>>selected spatial symmetry group.

>>It is impossible at present to solve for only a subset of symmetry types. You can solve for either all types (with no symmetry) or for single specific type.

>>All the best,
>>Alex
>>
>>
>>On Tue Jun 19 '12 3:29am, Solntsev Pasha wrote
>>----------------------------------------------
>>>Dear Alex.

>>>I working on CASSCF/GUGA. C4v symmetry.

>>>Could you explain negative value for ISTSYM? I found, for $drt group=c4v $end i can setup 0,1,2,3,4 values for istsym. Also i can setup -1,-2, -4 -5, -6. -3 doesn't work. In general, is it possible to request two different states. For example A1 and E? Thus, for istsym=0, istsym=1 and istsym=3 two possible states will be considered: A2/B2, A1/B1, A2/B2.

>>>Thank you very much.
>>>
>>>
>>>Pavel.
>>>
>>>
>>>