[molpro-user] confused about nstati
H. -J. Werner
werner at theochem.uni-stuttgart.de
Thu Jun 10 09:00:20 BST 2004
Apparently, at the shorter distances the order of the states in mrci is different
than in the casscf reference function, and this cases the error. This can happen
in the vicinity of avoided crossings. Increasing the number of states or increasing
nstati option can only help if all the states under consideration can be described
in the space of reference configurations. In the present case, it seems that there
is an additional state (perhaps a singly excited Rydberg state) which cannot be
described by the reference space (you would need more orbitals in the reference
wavefunction). You can either reduce the number of states (e.g., state,6) to exclude
this higher state, or increase the active orbital space by adding more orbitals.
However, with so many states it is often very difficult to find a set of orbitals
and states which yield smooth potentials for all geometries.
On Do, 10 Jun 2004, Wang Mingwei wrote:
>I am doing a CI calculation at different bonding length. when the r=1.667 or
>longer the job was done well. But from r=1.617 it always print such error
>NO VECTOR OF INTERNAL CI OVERLAPS SUFFICIENTLY WITH REFERENCE VECTOR 7.
>and the corresponding input file is followed here:
>But after I increased the nstati=12, I get such error info:
> ? Error
> ? nm.lt.n
> ? The problem occurs in rsjd
> ERROR EXIT
> CURRENT STACK: CIPRO MAIN
>What is more confusing is that when r=1.567 the calculation run well.
>Looking forward to your suggestion!
>College of Chemistry and Molecular Engineering
>mingwei at hpsv.pku.edu.cn
Prof. Hans-Joachim Werner
Institute for Theoretical Chemistry
University of Stuttgart
D-70569 Stuttgart, Germany
Tel.: (0049) 711 / 685 4400
Fax.: (0049) 711 / 685 4442
e-mail: werner at theochem.uni-stuttgart.de
More information about the Molpro-user