[molpro-user] How to restart CCSD(T) jobs?

[S.Manogaran]

I am new to molpro. So this is only a suggestion. 
I think this one works correctly.
First job, gets killed when cpu limit is reached.  The final
parameters are in the file h2o.act which is read in the
next job. I am not sure whether the int file is needed.

***,h2o
memory,400,m
file,1,h2o.int,new
file,2,h2o.wfu,new
file,3,h2o.aux,new
r=0.95782,theta=104.485          !set geometry parameters
geometry={ang
           O;              !z-matrix geometry input
           H1,O,r;
           H2,O,r,H1,theta}
basis=cc-pcvtz                 !use VTZ basis
hf;                       !closed-shell scf;
{ccsd(t);core,0}
{optg,saveact=h2o,savegrd=h2o,append};
put
---
***,h2o
memory,400,m
file,1,h2o.int
file,2,h2o.wfu
file,3,h2o.aux
geometry={ang
           O;              !z-matrix geometry input
           H1,O,r;
           H2,O,r,H1,theta}
readvar,h2o,noindex;
goto,optg;
basis=cc-pcvtz                 !use VTZ basis
hf;                       !closed-shell scf;
{ccsd(t);core,0}
{optg,saveact=h2o,savegrd=h2o,append};
put
---
S.Manogaran



On Fri, 13 Apr 2012, Vera Cathrine wrote:

> Dear Everybody,

I am running some UCCSD(T) computations. Our cluster wall-time is not enough to terminate the jobs normally. So I like to run them in a way to be able follow my job by restart from  the last UCCSD iteration and saving the time. I have specified the file 1 and 2 as following:


file,1,a.int,new
file,2,a.wfn,new

>From the manual I understand that I have to remove the new to restart the job. But it seems that this is not enough for CCSD(T) jobs and they are running from the scratch. What are the other directives which I have to specify to be able to use RESTART directive for rerunning my jobs?

Thanks in advance for your help.

Best wishes,
Vera