From owner-chemistry@ccl.net Wed Dec 15 10:34:00 2010 From: "=?utf-8?q?=C3=96d=C3=B6n_Farkas?= farkas]^[chem.elte.hu" To: CCL Subject: CCL: Schematic optimization sequence Message-Id: <-43395-101215052829-4566-ZcCfd+e19qbI/+1hoCwKaA^^^server.ccl.net> X-Original-From: =?utf-8?q?=C3=96d=C3=B6n_Farkas?= Content-Disposition: inline Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset="utf-8" Date: Wed, 15 Dec 2010 11:28:07 +0100 MIME-Version: 1.0 Sent to CCL by: =?utf-8?q?=C3=96d=C3=B6n_Farkas?= [farkas^^^chem.elte.hu] Hi Peter, People (and it seems zou also) mix the geometry optimization and constructing the wavefunction for the molecule. Finding the appropriate wavefunction is many times a variational problem, therefore itself contains some sort of optimization. However, what you refer to as Berny optimization (Named after its "father" prof. H. Bernhard Schlegel) is geometry optimization and its goal to find a stationary point on the potential energy (hyper)surface, which is defined by the wavefunction and the total energy of the system. You can find his papers on his webpage: http://www.chem.wayne.edu/schlegel/Current/Schlegel.html Unfortunately, other than the geometry optimization uses the gradient (first derivative of the energy with respect to atomic coordinates) and then predicts iteratively the position of the critical point until convergence (based on gradient and step size based limits) cannot be said on non-mathematical way about the process. If you have problems about the math involved I can help you but I cannot really gide you without involving at least some of the math concerned. I hope I could help. Best wishes, Ödön Odon Farkas Associate Professor Laboratory of Chemical Informatics Institute of Chemistry Eötvös Loránd University, Budapest 1/A Pázmány Péter sétány H-1117 Budapest, Hungary Cellphone: +36-30-2553111 On Monday 13 December 2010 13:28:37 P.D.Jarowski##surrey.ac.uk wrote: > Hi, thanks to all for their helpful comments on previous posts! > > I would like to know if there are any online sources out there that > describe in a didactic non-mathematical way how various computational ab > initio sequences are caried out. Especially, for multireference methods. > For example, what exactly in Berney optimization and how does is work > within the scf procedure? I would love to have such a source for my > students. > > Thanks, > > Best, > > Peter > ***Sent via RoadSync® for Android™