[molpro-user] SURF: printing mode # during 1D calc
Benj FitzPatrick
benjfitz at uchicago.edu
Thu Sep 3 23:42:16 BST 2009
Hello,
The radical (doublet) I'm trying to study is fairly troublesome in that I get
rohf convergence errors in the first 1D mode SURF tries (calculating the grid
points for mode 20). It has an epoxide ring that leads to rohf convergence
issues when the ring starts to open. I wanted to see if I could get a rough
estimate of the VCI modes by scaling back the grid calculations (using SCALE or
SCALNM) along the modes that lead to ring opening. Is there a way to get MOLPRO
to output the current mode (either in the .out or molpro.log) so that when it
crashes after a convergence problem I know which mode to scale? Thus far I've
been doing it the hard way by using SCALNM for a mode, seeing if the job
crashes, and, if so, choosing another mode.
Also, does SURF start calculating the grid at one of the end points and then
work towards the minimum and out to the other endpoint? This seems to be the
case given that changing scalnm solves the convergence problem. I was wondering
if I could get MOLPRO to start at the minimum and work out towards the ends,
which might avoid some of these convergence problems b/c the changes in the
orbitals will be more gradual from one grid point to the next.
Thanks,
Benj FitzPatrick
University of Chicago
More information about the Molpro-user
mailing list