Firefly and PC GAMESS-related discussion club

Learn how to ask questions correctly

sanya
sanya@photonics.ru

Thank you!

>>What off-diagonal value should be considered as large: ~0.1, or ~0.01, or ~1?
>I'd suggest to avoid off-diagonals larger than ca. 0.1-0.2
>and safely ignore smaller values. Very large off-diagonal values
>of overlap of two XMCQDPT2 states typically mean that these two
>states attempt to describe the same real state. The most
>likely reason of this is the poor choice of the model space
>(e.g., due to some low-lying spurious CASSCF states entering
>model space). Another reason could be intruder state problem.
>> and is it safe if the imaginary part of an eigenvalue of non-symmetric H_eff is ~1.0e-16?
>It is absolutely safe. These small values are artifacts of
>accumulation of roundoff errors inside non-symmetric matrix
>diagonalization code. Look at the output and you'll find that
>complex eigenvalues are always coming in pairs of conjugated
>values.

That's nice!

>> And what should I do in case of problems? change some XMCQDPT
>> parameters or start with another CASSCF solution?
>In most cases this means you need to double-check CAS active
>space and CASSCF solutions.

I thought so.

>>>>By the way, in the next FF version, is it possible to print the WARNING that off-diagonal values are large or some eigenvalues of non-symmetric H_eff are complex?
>>>Is it too difficult to check this manually in the output? :)
>>For me, it is not difficult, but there are many less experienced FF users who knows nothing about possible problems. And I learned about this issue just now, in this discussion. Well, it may follow from your paper in J. Chem. Phys., but it is still in press and not available...

>The paper will be available online in a week or so.

It's wonderful news :)

>This problem is not discussed in the paper. Actually, these are the
>well-known facts for a (very) small audience. :)

So, it is even more important to make them accessible for a broad audience ;)