[ghemical-users] Ghemical problems

[Jean Bréfort]

Le vendredi 09 avril 2010 à 09:27 +0300, Tommi Hassinen a écrit :

I'm using mopac7 1.15 and mpqc 2.3.1

Using the default qm method (MOPAC7 / MNDO) I get a crash with a methane
molecule:

0  0x00007fffeb42af73 in lm7_ini_full_xyz () from /usr/lib/libmopac7.so.1
#1  0x00007ffff36dfa95 in eng1_qm_mopac::eng1_qm_mopac(setup*, unsigned int, unsigned int) () from /usr/lib/libghemical.so.5
#2  0x00007ffff36d573f in setup1_qm::CreateEngineByIndex(unsigned int) ()
   from /usr/lib/libghemical.so.5
#3  0x00007ffff36b1515 in setup::CreateCurrentEngine() ()
   from /usr/lib/libghemical.so.5
#4  0x00007ffff36ae56f in model::DoEnergy() () from /usr/lib/libghemical.so.5

Trying to do the same with just a He atom makes the program exit, but
with no crash, according to gdb.

Jean

> Hello,
> 
> can you send me the mopac7 and MPQC versions you are using?
> 
> About documentation, you can type a command "help" in the "Command Interpreter"
> command string dialog box to get a brief listing of available commands and parameters
> related to them. I don't know any other documentation (at least I haven't written any).
> This project suffers from a serious lack of resources now, for example I am not working
> in a university position right now. I can't write documentation now (must pay the bills you
> know) but I will maintain Ghemical so that it keeps working and is available anyway.
> 
> About QM geometry optmization in Ghemical, it works so that it only asks energy+force
> information from the QM package and uses an optimization algorithm of it's own. The
> optimization algorithm is best tuned for MM problems, so it is possible that the QM
> package itself can do the optmization quicker.
> 
> About plotting molecular orbitals, you should use the "Set Current Orbital" popup menu
> item to select which MO you wish to plot ; you can use the
> 
> Render / Views / Create / Energy-level Diagram
> 
> view to see the possible MO:s for your molecule. The whole QM system in Ghemical
> currently works only for "closed shell" molecules (spin multiplicity can only be 1).
> 
> Regards,
> 
> Tommi
> 
> ________________________________________
> From: ghemical-users-bounces at bioinformatics.org [ghemical-users-bounces at bioinformatics.org] On Behalf Of Glen Shennan [glen.shennan at gmail.com]
> Sent: Friday, April 09, 2010 1:16 AM
> To: ghemical-users at bioinformatics.org
> Subject: Re: [ghemical-users] Ghemical problems
> 
> Turns out I shouldn't have been trying ab initio QC on my laptop.  I was trying to model thiophene when the crashes occured but when I try ethane I can get through the energy calculation and geometry optimisation (slowly) with no crashes, using MPQC STO-3G.  Assume lack of memory is to blame?
> 
> I want to use ghemical/mpqc to produce images of individual molecular orbitals in organic species but I am really struggling to find documentation.  Is there a reference that would get me up to speed?  I understand this is possible with ghemical but I might be wrong?
> 
> Thanks.
> 
> On 9 April 2010 05:16, Jean Bréfort <jean.brefort at normalesup.org<mailto:jean.brefort at normalesup.org>> wrote:
> Getting a crash as well on debian.
> 
> _______________________________________________
> ghemical-users mailing list
> ghemical-users at bioinformatics.org
> http://www.bioinformatics.org/mailman/listinfo/ghemical-users
>