[molpro-user] save/start,task in frequency calculations

[Jacky LIEVIN]

Dear all,

I'm playing with the "masktask" option (together with "save,task=record number" and "start,task=record number" ) in order to split in several jobs a time consuming numerical frequency calculation with MRCI.

Task restarting is working fine for the lowest state of a multi-state MRCI calculation, but unfortunately it fails for the excited states: the gradient norm is recalculated as not zero at equilibrium position and the hessian and resulting frequencies are wrong.  Log files with and without restart are yet identical for all tasks. I thus guess that the problem occurs in the numerical hessian calculation. Why?

Maybe that some flag (not reported in the manual) should be added in the save/start task directive?
I tried  … save,task=5500.2,state=2.1  … not rejected by the program … but giving the same incorrect results

Any help, example input would be welcomed
Many thanks in advance

Jacky

_____________________________________
Prof. Jacky Liévin
Service de Chimie Quantique et Photophysique
Université Libre de Bruxelles, CPi 160/09
50 av F.D. Roosevelt
B-1050 Bruxelles
Belgium
Tel: +32-2-650 4089
Fax: +32-2-650 4232
_____________________________________