[molpro-user] Segmentation fault when computing MRCI transition dipoles

[Jacky LIEVIN]

dear Sasha,

I'm adressing this mail to you because you already answered before to the same problem on the mailing list (see posted mails below)
I'm getting a "segmentation violation" error in the calculation of transition dipoles with the "biorth" option (compiler is ifort).

I'm wondering if the modified code you mentioned in july has now be included in the recent versions of molpro (I'm using molpro2012.24)?
I also tried the suggestion you made on the mailing list, which is to compile mrci/citra_conf.F using -O1, but this doesn't fix the problem
In the mean time I use an old 2010 version compiled with pgf90, which works fine for biorth calculations

The following input can be used for testing:
***,X and A states of H2O
gthresh,energy=1.d-8
r=0.957,angstrom,theta=104.6,degree;
geometry={O;              
          H1,O,r;
          H2,O,r,H1,theta}
{hf;wf,10,1;}
{multi;occ,4,1,2;closed,2;wf,9,2,1}
{ci;occ,4,1,2;closed,2;core,1;wf,9,2,1;save,7300.1}
{multi;occ,4,1,2;closed,2;wf,9,1,1}
{ci;occ,4,1,2;closed,2;core,1;wf,9,1,1;save,7100.1}
{ci;trans,7300.1,7100.1,biorth}


Thanks in advance for any help

Jacky





Le 20 juil. 2014 à 12:42, Alexander Mitrushchenkov <Alexander.Mitrushchenkov at u-pem.fr> a écrit :

> Lorenzo,
> 
> Finally I have simplified a code and now compiler is happy. If you use development version you can pull right now; otherwise wait a bit until it will be backported to 2012. Or I could send modified citra_conf.F directly to you if you wish.
> 
> Best,
> 
> Sasha
> 
> Le 08/07/2014 16:01, Lorenzo Lodi a écrit :
>> I'm trying to compute electronic transition dipoles with MRCI with
>> non-orthogonal orbitals and I'm getting a `segmentation fault' error at
>> runtime.
>> 
>> For example, consider the following example for NaH:
>> 
>> basis={default, cc-pVDZ}
>> r = 3.5
>> geometry={Na; H, Na, r}
>> {MCSCF; OCC,6,2,2,0; CLOSED,3,1,1,0; FROZEN,0,0,0,0; WF,
>> elec=12,sym=1,spin=0; STATE,1; LQUANT,0}
>> {MRCI ; OCC,6,2,2,0; CLOSED,3,1,1,0; CORE,3,1,1,0  ; WF,
>> elec=12,sym=1,spin=0; STATE,1; SAVE, 3001.1}
>> {MCSCF; OCC,6,2,2,0; CLOSED,3,1,1,0; FROZEN,0,0,0,0; WF,
>> elec=12,sym=1,spin=0; STATE,2; LQUANT,0,0; WEIGHT,0.1, 1}
>> {MRCI ; OCC,6,2,2,0; CLOSED,3,1,1,0; CORE  ,3,1,1,0; WF,
>> elec=12,sym=1,spin=0; STATE,1,2; SAVE, 3101.1}
>> {MRCI; TRANS, 3001.1, 3101.1, BIORTH,DMZ}
>> 
>> The idea is to compute DMZ between the ground 1Sigma+ state and the
>> first excited 1Sigma+ state, with each state using its own optimized
>> MCSCF orbitals.
>> When I try to run it with my installation of molpro 2012 I get:
>> 
>> 
>> 1PROGRAM * CI (Multireference internally contracted CI)     Authors:
>> H.-J. Werner, P.J. Knowles, 1987
>> 
>> 
>>    Transition moment calculation
>>    =============================
>> 
>> Received signal 11 Segmentation violation
>>  =============================================================================
>>  |  Orbitals are different, transformation to biorthogonal orbitals is
>> used |
>>  =============================================================================
>>  1PROGRAM * CITRA (MRCI wavefunction orbital transformation)  Author :
>> A.O.Mitrushenkov, 2005
>> 
>>  fehler -1701337647 (0x9a97a5d1).
>>  fehler -1701337647 (0x9a97a5d1).
>>  PPIDD_Error: now exiting...
>> 
>> 
>> 
>> Any hints as to what may be causing the problem? Is the input correct?
>> If I use state averaged MCSCF orbitals for both MRCI runs then I get the
>> correct result (no crash).
>> 
>> Lorenzo
>> 
>> _______________________________________________
>> Molpro-user mailing list
>> Molpro-user at molpro.net
>> http://www.molpro.net/mailman/listinfo/molpro-user
> 
> 
> -- 
> Dr. Alexander Mitrushchenkov, IGR
> Laboratoire de Modélisation et Simulation Multi Echelle
> UMR 8208 CNRS
> Université Paris-Est Marne-la-Vallée
> 5 Bd Descartes
> 77454 Marne la Vallée, Cedex 2, France
> 
> Phone:    +33(0)160957316
> Fax:      +33(0)160957320
> e-mail:   Alexander.Mitrushchenkov at u-pem.fr
> 
> 
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user

_____________________________________
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
_____________________________________