From jaouad "-at-" csc.fi Tue Jun 9 05:29:45 1998 Received: from pobox.csc.fi (pobox.csc.fi [128.214.46.62]) by www.ccl.net (8.8.3/8.8.6/OSC/CCL 1.0) with ESMTP id FAA01084 Tue, 9 Jun 1998 05:29:42 -0400 (EDT) Received: from cypress.csc.fi (cypress-e.csc.fi [128.214.46.7]) by pobox.csc.fi (8.9.0.Beta5/8.9.0.Beta5/CSC/Rtrs-1.22) with SMTP id HAA05888; Tue, 9 Jun 1998 07:13:20 +0300 (EET DST) Date: Tue, 9 Jun 1998 07:12:43 +0300 (EET DST) From: Jaouad El-Bahraoui Reply-To: Jaouad El-Bahraoui To: Melissa Petersen cc: chemistry "-at-" www.ccl.net Subject: Re: Gaussain 94 Error Message In-Reply-To: <9806051821.ZM4742- at -sg10.chem.tue.nl> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi Melissa and CCl's A while ago I asked the below question: Does any one know how to solve this kind of problem in g94, I use the version E.2 Density matrix breaks symmetry, PCut= 1.00D-04 Density matrix has no symmetry -- integrals replicated. Density matrix breaks symmetry, PCut= 1.00D-04 Density matrix has no symmetry -- integrals replicated. Here I present the summary of responses, and by the way I want to thanks all who gives me a little of his time to solve my problem. all the Best Jaouad *************************************************************************** summary *************************************************************************** On Fri, 5 Jun 1998, Melissa Petersen wrote: > Hi Jaouad, > > I have just joined the computational chemistry list and I noticed that you > posted a message on 25 May concerning an error that you got in g94 i.e. > "Density Matrix breaks symmetry" > > I have also been experiencing this problem. Could you tell me if you have > received any replies? I was not able to find any replies in the message > archives. > > Thanks in anticipation. > Melissa Petersen > > ******************************************************************************** > ******************************************************************************** > Catalysis Research Unit > Department of Chemical Engineering > University of Cape Town > Private Bag, Rondebosch, 7701 > Cape Town > South Africa > > Email: melissa*- at -*chemeng.uct.ac.za > or tgakmp /at\sg10.chem.tue.nl > **************************************************************************************************************************************************************** >From hofmann |-at-| thkin.pci.uni-leipzig.de Tue Jun 9 06:53:38 1998 Date: Mon, 25 May 1998 16:23:42 +0100 (GDT) From: hofmann -8 at 8- thkin.pci.uni-leipzig.de To: Jaouad El-Bahraoui Subject: Re: CCL:G:G94 problem..... Hi! Your problem may be a numerical one. Density matrix breaks symmetry, PCut= 1.00D-04 Density matrix has no symmetry -- integrals replicated. Density matrix breaks symmetry, PCut= 1.00D-04 Density matrix has no symmetry -- integrals replicated. It arises if the integrals must be the same in a certain symmetry, but they are not. So one hint is to tell the scf-procedure to calculate integrals with respect to symmetry: scf=(intrep) This can be keyword connected to scf=(symm). This one frozens configurations and is a powerful tool to get crazy things. Be careful. Alex ------------------------------------------------------------------------- | | | | Alexander Hofmann | Phone (office) +49-341-97-36337 | | Wilhelm-Ostwald-Institut fuer | Fax (office) +49-341-97-30099 | | Physikalische und Theoretische | | | Chemie | | | Augustusplatz 10-11 | hofmann-!at!-thkin.pci.uni-leipzig.de | | | | | D-04109 Leipzig | http://www.uni-leipzig.de/~quant | | | /hofmann/alex.html | ------------------------------------------------------------------------- >From danne # - at - # rscqc2.anu.edu.au Tue Jun 9 06:53:49 1998 Date: Tue, 26 May 1998 01:42:04 +1000 From: Danne R Rasmussen To: Jaouad El-Bahraoui Subject: Re: CCL:G:G94 problem..... a few options: a) if you are reading in a guess (using guess=read or guess=check) then don't, b) try using SCF=novaracc i.e. use best integral accuracy at all steps in the SCF, c) if you are doing DFT try SCF=vshift and/or SCF=(vshift,novaracc), d) if you are using many diffuse functions in a large molecule try using a different basis set; diffuse functions can cause convergence problems due to linear dependence etc, e) (risky) if you think you should have symmetry but keep running into problems (i.e. all the above failed) try modifying PCut in the G94 source and recompiling l502 [PCut=1.0D-03 will do it but you risk getting false symmetry]. you can put your special l502.exe in your own home directory (e.g. ~/g94/l502.exe) and modify GAUSS_EXEDIR to include ~/g94 earlier in the path than the $root/g94 directory then your own l502.exe will be used instead of the default executable. good luck, danne >From Krys.Radacki -A_T- ac.rwth-aachen.de Tue Jun 9 06:53:55 1998 Date: Mon, 25 May 1998 21:11:30 +0200 From: Krzysztof Radacki To: Jaouad El-Bahraoui Subject: Re: CCL:G:G94 problem..... As far as I know it means that during geometry optimization you have lost symmetry. Try to start without symmetry (NoSymm flag in command line) for example: #P RHF/6-31++G(d,p) opt=z-matrix Nosymm regards Krzys Radacki >From b_duke ^at^ quoll.ntu.edu.au Tue Jun 9 06:54:00 1998 Date: Tue, 26 May 1998 09:09:21 +0930 (CST) From: Brian Salter-Duke To: Jaouad El-Bahraoui Subject: Re: CCL:G:G94 problem..... Hi, 1. Check everything. Have you got the right state you want? 2. Do you have an ion? If so do the parent closed shell molecule and use guess=read to get the starting vectors. Maybe try reading the coefficients from a smaller basis set. Cheers, Brian. -- Associate Professor Brian Salter-Duke (Brian Duke) School of Mathematical and Physical Sciences, Northern Territory University, Darwin, NT 0909, Australia. Phone (61) (0)8-89466702 e-mail: b_duke[ AT ]lacebark.ntu.edu.au or b_duke[ AT ]quoll.ntu.edu.au >From Megan.Govender /at\eskom.co.za Tue Jun 9 06:54:05 1998 Date: Tue, 26 May 1998 06:48:54 +0200 From: Megan Govender To: jaouad-: at :-csc.fi, jaouad-: at :-csc.fr Subject: Re: CCL:G:G94 problem..... Hi.... I also had that same problem optimizing compounds of Uranium some time back... Are you trying similar calculations?...I used to use opt=fp with the nosymm option for these type calculations to get around this problem. Hope this helps... Regards Magan >From simek ":at:" rudjer.irb.hr Tue Jun 9 06:54:20 1998 Date: Wed, 27 May 1998 12:40:11 +0200 From: Visnja Simek To: Jaouad El-Bahraoui Subject: Re: CCL:G:G94 problem..... Do you put "NoSym" in # line in addition to other keywords ? Best regards, Visnja ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Visnja Simek Laboratory of Molecular Spectroscopy Rudjer Boskovic Institute P.O.B. 1016, Zagreb CROATIA Tel: (385-1) 456 09 44 Fax: (385-1) 468 00 84 (385-1) 468 01 95 E-mail: simek-!at!-rudjer.irb.hr ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - \^ ^/ ) %-% at %-% %-% at %-% ( +-----------------------oOO--(_)----------------------------------+ | Jaouad El-Bahraoui , Ph.D | +-----------------------------------------------------------------+ | Inorganic Chemistry Laboratory | Tel: +358-9-19140199 | | department of Chemistry, P.O. BOX 55 | Fax: +358-9-19140198 | | University of Helsinki,FIN-00040 | jaouad#* at *#nd.edu | | Helsinki, Finland | w3.nd.edu/~jaouad/ | +----------------------Oooo--oOO----------------------------------+ oooO ( ) ( ) ) / \ ( (_/ \_) -------------------------------------------------------------------- I know you believe you understand what you think I said, but I am not sure you realize that what you heard is not what I meant. --------------------------------------------------------------------