[molpro-user] UMP2/uhf for natrual orbitals

[Hans-Joachim Werner]

See manual, section "16.1.5 Special options for UHF calculations". If UHF  natural orbitals are used
as starting guess for multi, it determines the active space using the occupation number thresholds UNOMIN, UNOMAX.
This happens by default if multi follows an UHF calculation.
This behaviour can be overwritten by the usual „occ" and „closed" directives. UMP2 is a poor method if
spin-contamination is strong and should in my opinion not be used. Unfortunately, we do currently not
have uccsd based on uhf orbitals. 
Best regards
Joachim Werner
---
Prof. Dr. Hans-Joachim Werner
Institut for Theoretical Chemistry
University of Stuttgart
Pfaffenwaldring 55
D-70569 Stuttgart
Tel: +49 711 / 685 64400
Fax: +49 711 / 685 64442
email: werner at theochem.uni-stuttgart.de



Am 17.03.2014 um 08:47 schrieb Bo Long <wwwltcommon at sina.com>:

> Dear Prof Werner:
>   I am so sorry to interrupt you. I face some situations in which wavefunction can not be well described using rmp2 based on rhf. I want to do UMP2 based on UHF. I do not know whether CASSCF calculation can read natural orbital from the previous UMP2/UHF? Thank you for your help.
> Best Wishes
> Bo Long