In the first case, the program uses Slater determinants as a basis, which is the default. In this case, (-) means beta spin and (+) alpha spin. If you save the reference vector for later use in mrci, CSFs have to be used, and the program automatically switches to this (same as giving the CONFIG directive). In this case, a genealogical spin coupling scheme is used, and +/- corresponds to going up or down in the branching diagram (to be read from left to right). Joachim Werner ----- Forwarded message from Tim Robinson <[EMAIL PROTECTED]> -----
From: Tim Robinson <[EMAIL PROTECTED]> To: [EMAIL PROTECTED] Subject: [molpro-user] CI vector: the role of the observer (?) Date: Thu, 10 Jun 2004 15:12:34 +1200 (NZST) Message-ID: <[EMAIL PROTECTED]> Hi all, I see different behaviour in the CI vectors, depending on whether I save the vector or don't save it. For example, a simple CASSCF calculation on formaldehyde produces the following CI vector info (print threshold at 0.04) CI vector --------- 20 20 20 0.9769517 20 02 20 -0.1360811 02 20 20 -0.0560268 20 20 02 -0.0548364 -+ +- 20 -0.0523983 +- -+ 20 -0.0523983 20 -+ +- -0.0459565 20 +- -+ -0.0459565 TOTAL ENERGIES -113.98318578 ... whereas if I say I want to save the CI vector (multi;save,ref=XXXX.Y) then the output becomes: CI vector --------- 20 20 20 0.9769517 20 02 20 -0.1360810 +- +- 20 0.0863966 20 +- +- 0.0807951 +- 20 +- -0.0586269 02 20 20 -0.0560267 20 20 02 -0.0548366 TOTAL ENERGIES -113.98318578 In particular, the 5th configuration in the second example doesn't even (seem to) appear in the first output at all. Perhaps I don't properly understand how these CI vectors are to be read. Is it something to do with CSFs vs Slater dets? Confused, Best regards Tim ---- Tim Robinson 5n9 Department of Chemistry University of Otago 64 3 479 5214 ----- End forwarded message ----- -- Prof. Hans-Joachim Werner Institute for Theoretical Chemistry University of Stuttgart Pfaffenwaldring 55 D-70569 Stuttgart, Germany Tel.: (0049) 711 / 685 4400 Fax.: (0049) 711 / 685 4442 e-mail: [EMAIL PROTECTED]
