From chemistry-request@server.ccl.net Wed Feb 23 02:08:43 2000 Received: from gateway.tbzmed.ac.ir ([195.200.227.150]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id CAA10743 for ; Wed, 23 Feb 2000 02:08:31 -0500 Received: from CLIENT2 ([192.168.128.212]) by gateway.tbzmed.ac.ir with SMTP (Microsoft Exchange Internet Mail Service Version 5.5.2650.21) id FNZMV71T; Wed, 23 Feb 2000 09:38:04 +0330 Message-ID: <000501bf7dc2$3a577200$d480a8c0@res.tbzmed.ac.ir> From: "Dr. B. Habibi-Nazhad" To: Subject: Mutation Database ? Date: Wed, 23 Feb 2000 09:22:37 +0330 Dear Netters; I would be so thankful if you kindly inform me about Web address(s) of suitable Mutation Database(s) for proteins. Thank you. B. Habibi-Nazhad, Ph.D **************************************************************************** Department of Medicinal Chemistry, Faculty of Pharmacy, Tabriz University of Medical Sciences, ZIP Code: 51657 Tabriz, Iran. **************************************************************************** From chemistry-request@server.ccl.net Wed Feb 23 03:48:24 2000 Received: from gate-sl1.mdli.com (ns2.mdli.com [208.200.221.3]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id DAA11168 for ; Wed, 23 Feb 2000 03:47:51 -0500 Received: (from smap@localhost) by gate-sl1.mdli.com (8.8.8/8.8.8) id XAA24990 for ; Tue, 22 Feb 2000 23:40:44 -0800 (PST) Received: from puffin.mdli.com(191.254.19.10) by gate-sl1.mdli.com via smap (V2.1) id xma024988; Tue, 22 Feb 00 23:40:40 -0800 Received: from shepherd.mdli.com by puffin.mdli.com (980427.SGI.8.8.8/BCH1.0) id XAA07829; Tue, 22 Feb 2000 23:40:40 -0800 (PST) Received: from mdli.com (lchen-95h.mdli.com [191.254.19.100]) by shepherd.mdli.com with SMTP (Microsoft Exchange Internet Mail Service Version 5.5.2650.21) id FK4MG5S6; Tue, 22 Feb 2000 23:40:40 -0800 Message-ID: <38B38C76.C0CE0DC8@mdli.com> Date: Tue, 22 Feb 2000 23:29:58 -0800 From: Lingran Chen Organization: MDL Information Systems, Inc. X-Mailer: Mozilla 4.7 [en] (WinNT; I) X-Accept-Language: en MIME-Version: 1.0 To: CHEMISTRY@ccl.net Subject: The "birth-year" of computational chemistry? Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Dear CCLers: In the Encyclopedia of Computational Chemistry (ECC), there are some interesting statements about the "birth-year" of computational chemistry. For example, 1. "Computational chemistry has existed for half a centrury, ..." - Forword of ECC by Prof. Pople (1997). Thus, the birth-year of CC is about 1947. 2. "Although its origin date back much earlier, Computational Chemistry evolved into a distinctly discernible discipline three decades ago ..." - Preface of ECC by Prof. Schleyer (1997?) Thus, the birth-year of CC is about 1967. 3. "The general availability of computers at major research laboratories and the availability of FORTRAN as a general programming language to make use of these computers marks 1960 as approximately the beginning of modern computational chemistry ..." - "HISTORY OF COMPUTATIONAL CHEMISTRY: A PERSONAL VIEW" by prof. Streitwieser. (ECC, p.1238) Here, the birth-year of CC is 1960. I have two questions: 1. Who published the first paper about using an electronic computer to solve a(ny) chemical problem? In my personal record, the earlest publication is: P D. Zemany, "Punched Card Catalog of Mass Spectra Useful in Qualitative Analysis," Anal. Chem., 22, 1950, 920. 2. Who wrote the first quantum chemistry program? The answers to the above questions may indicate the "reasonable" birth-year of CC. Thanks in advance. -Lingran *********************************** Lingran Chen, Ph.D. Advisory Scientist MDL Information Systems, Inc. 14600 Catalina Street San Leandro CA 94577 U.S.A. Phone: (510) 895-1313 FAX: (510) 614-3616 Email: LCHEN@MDLI.COM Web: http://www.mdli.com *********************************** From chemistry-request@server.ccl.net Wed Feb 23 04:59:38 2000 Received: from sunu450.rz.ruhr-uni-bochum.de (sunu450.rz.ruhr-uni-bochum.de [134.147.64.5]) by server.ccl.net (8.8.7/8.8.7) with SMTP id EAA11664 for ; Wed, 23 Feb 2000 04:59:38 -0500 Received: (qmail 12053 invoked from network); 23 Feb 2000 08:52:35 -0000 Received: from cvw.orch.ruhr-uni-bochum.de (HELO ?134.147.110.57?) (134.147.110.57) by mailhost.rz.ruhr-uni-bochum.de with SMTP; 23 Feb 2000 08:52:35 -0000 Mime-Version: 1.0 X-Sender: wuellcbd@mailhost.rz.ruhr-uni-bochum.de Message-Id: In-Reply-To: <38B38C76.C0CE0DC8@mdli.com> References: <38B38C76.C0CE0DC8@mdli.com> Date: Wed, 23 Feb 2000 09:52:31 +0100 To: Lingran Chen From: Christoph van =?iso-8859-1?Q?W=FCllen?= Subject: Re: CCL:The "birth-year" of computational chemistry? Cc: Chemistry@CCL.net Content-Type: text/plain; charset="us-ascii" ; format="flowed" > >1. Who published the first paper about using an electronic >computer to solve a(ny) chemical problem? > >In my personal record, the earlest publication is: > >P D. Zemany, "Punched Card Catalog of Mass Spectra Useful in Qualitative >Analysis," >Anal. Chem., 22, 1950, 920. > I have no reference at hand, but I think computers were used even before in the field of X-Ray structure analysis. -- ---------------------------+------------------------------------------------ Christoph van Wullen | Fon (University): +49 234 32 26485 Theoretical Chemistry | Fax (University): +49 234 32 14109 Ruhr-Universitaet | Fon/Fax (private): +49 234 33 22 75 D-44780 Bochum, Germany | eMail: Christoph.van.Wuellen@Ruhr-Uni-Bochum.de ---------------------------+------------------------------------------------ From chemistry-request@server.ccl.net Wed Feb 23 05:42:12 2000 Received: from main.amu.edu.pl (root@main.amu.edu.pl [150.254.65.7]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id FAA11904 for ; Wed, 23 Feb 2000 05:42:11 -0500 From: jolanala@main.amu.edu.pl Received: from zyxel (pa71.poznan.ppp.tpnet.pl [212.160.10.71]) by main.amu.edu.pl (8.9.1a/8.9.1) with SMTP id KAA00051 for ; Wed, 23 Feb 2000 10:31:50 +0100 (MET) Message-Id: <200002230931.KAA00051@main.amu.edu.pl> To: chemistry@ccl.net Date: Wed, 23 Feb 2000 10:27:36 +0100 MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Subject: chlorates X-Confirm-Reading-To: jolanala@main.amu.edu.pl X-pmrqc: 1 Priority: normal X-mailer: Pegasus Mail for Win32 (v3.01d) Content-Transfer-Encoding: 7BIT Dear Netters, Do you have any idea where can I buy chlorates of Sr, Mg, Cs, Cu, Zn? I check many pharmaceutical and chemical companies without any success. I really appreciate your help, Thanks in advance, Jolanta From chemistry-request@server.ccl.net Wed Feb 23 05:49:00 2000 Received: from ccl.net (atlantis.ccl.net [192.148.249.4]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id FAA11974 for ; Wed, 23 Feb 2000 05:49:00 -0500 Received: from lumiere.idris.fr (root@lumiere.idris.fr [130.84.8.14]) by ccl.net (8.8.6/8.8.6/OSC 1.1) with ESMTP id EAA02269 for ; Wed, 23 Feb 2000 04:41:52 -0500 (EST) Received: from mira.idris.fr (gondet@mira.idris.fr [130.84.12.100]) by lumiere.idris.fr (8.9.3/8.9.3) with ESMTP id KAA11751 for ; Wed, 23 Feb 2000 10:41:48 +0100 (MET) Received: from mira.idris.fr (gondet@localhost) by mira.idris.fr (8.9.3/8.9.3) with ESMTP id KAA95508 for ; Wed, 23 Feb 2000 10:41:44 +0100 Message-Id: <200002230941.KAA95508@mira.idris.fr> X-Mailer: exmh version 2.0.2 2/24/98 To: chemistry@ccl.net Subject: CCL:Summary of Random Number Generators LGGES-SEQMATH Date: Wed, 23 Feb 2000 10:41:44 +0100 From: Gondet Etienne Hello, The following is a summary of the replies to my question about random number generators. **************************************************************** The original question. Hello, I am starting to do Quantum Monte Carlo calculations and I am interested in the performance of random number generators. There have been a number of tests developed, the most trusted seem to be those in "Seminumerical Algorithms" by Knuth. And there are a number of supplemental random number generators, such as the ran3 generator in "Numerical Recipies". It has also been demonstrated that in the past the random number generators built into various programming languages were not to be trusted. My question is - Do the random number generators built into the current generation of compilers pass statistical tests? I have tested the built in random number generators in Borland C++ and GNU C++ and so far they have passed all tests (working out of Knuth). I would like to know if it is only C++ that has a reliable random number generator implemented or if all compilers are now reliable, or if it is only certain vendors even within the C++ language? The reason that I would like to use built in random number generators is that Knuth indicates that code written directly in assembly language can accomplish the task faster than code written in standard source code. The GNU C++ built in random number generator is a factor of 2 faster than the ran3 generator using all available compiler optimization. **************************************************************** From: TOPPER ROBERT My best advice is to beware of using random number generators as black-boxes, especially in the context of Monte Carlo integrations and simulations. When using random number generators in Monte Carlo, you have to not only worry about whether the samples are uniform on the plane and uncorrelated in hyperdimensions but also about the period of the random number sequence (all random number generators, as you know, have a period for a given seed - and if your program requires a # of random numbers which is on the order of the period of the algorithm, better watch out). What's more, RNGs which perform perfectly well against a battery of statistical tests have been known to fail when used, for example, in 3D Ising lattice calculations...there is a good bit of literature on this subject. I personally wouldn't use any random number generator in my Monte Carlo work unless I knew exactly what the algorithm was, had a reference for it, and had subjected it to my personal battery of tests, which includes statistical tests ala Knuth and application tests (path-integral and classical Monte Carlo computations). At least then, if there is a bug in the algorithm I know who to blame for all those wasted cycles... Finally, in every Monte Carlo calculation I have ever carried out, the limiting factor has not been the speed of the random number generator, but the evaluation of the potential energy. But then I've never done QMC calculations, so I don't know what the "rate-determining step" is for such calcs. Because of this, I tend to err on the side of slow, but solid random number generation. hope this helps, best to all- rqt ************************************************************************ From: "Trevor Creamer" Dave, What you want to do is use the most reliable (statistically) random number generator available. I haven't done any Quantum Monte Carlo, but I've a lot of experience using MC to study peptides and polyelectrolytes. The fraction of time in a simulation spent generating the random numbers is miniscule compared to the time spent calculating the energies. It is essential that you use a reliable generator, much more important than using the fastest generator. You probably want to spend your time optimising your energy calculation code - this is where the bulk of simulation time is spent. Cheers, Trevor **************************************************************** From: hinsenk@ERE.UMontreal.CA (Hinsen Konrad) My question is - Do the random number generators built into the current generation of compilers pass statistical tests? In general, I doubt that. Random number generators are usually provided for games and other less demanding applications. The exception is the set of generators in the GNU C++ library. Others might also be good, but I would always check them first. (working out of Knuth). I would like to know if it is only C++ that has a reliable random number generator implemented or if all compilers are now reliable, or if it is only certain vendors even within the C++ language? I would assume the latter... The reason that I would like to use built in random number generators is that Knuth indicates that code written directly in assembly language can accomplish the task faster than code written in standard source code. The GNU C++ built in random number generator is a factor of 2 faster than the ran3 generator using all available compiler optimization. Nevertheless the GNU C++ random generators are written entirely in C++ (remember that GNU C++ is meant to be a portable compiler). **************************************************************** From: rs0thp@rohmhaas.com (Dr. Tom Pierce) Previously, Dave Young wrote: > I am starting to do Quantum Monte Carlo calculations and > I am interested in the performance of random number generators. > > There have been a number of tests developed, the most trusted > seem to be those in "Seminumerical Algorithms" by Knuth. And there > are a number of supplemental random number generators, such as the > ran3 generator in "Numerical Recipies". It has also been demonstrated > > My question is - Do the random number generators built into > the current generation of compilers pass statistical tests? > A simple answer is "No". I think this topic is a permanent flame-war in sci.math.stat. The current publications of interest are "Monte Carlo Simulations: Hidden errors from "Good" Random number generators" Phys. Rev. Letters 69(23), pp 3382 1992 Specifically this article says ".. a specific algorithm must be tested together with the random number generator being used REGARDLESS of the tests which the generator has passed." "Latest Development in Random Numbers", CERN Computer Newsletter 213, July-Sept. pp 5-7, 1993. and "Ultra, The greatest random number generator that ever was or ever will be", by Arif Zaman (arif@stat.fsu.edu) and George Marsaglia (geo@stat.fsu.edu), 1992. My painful experience with compiler random number generators is that one is far better off using a known and tested generator. **************************************************************** From: Gislason Jason I have used the random number generators found in "Numerical Recipes" and I have found them to work very well. Personally, I would not trust the generators built into compilers. You just don't know how good it works unless you or someone else has tested it. Jason Gislason **************************************************************** From: brianh@scg.scg.fujitsu.com (Brian Hammond) For QMC calculations the "quality" of pseudo-random numbers depends on the algorithm used. Most modern Fortran- or C-library pseudo-random number generators are good enough, particularly if you are doing branching in which you are using a random number of pseudo-random numbers. Also you must consider how to create Gaussian distributed random numbers if your algorithm calls for them. I have seen presentations where the overall cycle length of the underlying pseudo-random number generator is quite noticeable due to creating Gaussian numbers by adding 12 uniformly distributed numbers. So, in general use a pseudo-random number generator with a long cycle length and use Box-Muller if you want Gaussian pseudo-random numbers. -Brian **************************************************************** From: Markus.Wagener@EROS.CCC.Uni-Erlangen.DE Hello Dave, I always use the following random number generator. I might not be the fastest one, but it is portable and therefore gives the same results on all platforms. It is a linear congruential one. Additive random number generators, like those given in Knuth's book should be a lot faster! I hope tis helps! Markus <<<<<<<<<<<<<<<<<<<<<<<<<<<>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> #include #include /****************************************************************/ /* Rand computes a psuedo-random */ /* double value between 0 and 1, excluding 1. Randint */ /* gives an integer value between low and high inclusive. */ /****************************************************************/ #define Randomize() (Seed = time((time_t)0)) #define Randint(low,high) ( (int) ((low) + ((high)-(low)+1) * Rand())) #define Flip(P) ((P == 1.0) ? 1 : (Rand() <= P) ) extern double Rand(void); <<<<<<<<<<<<<<<<<<<<<<<<<<<>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> int Seed = 707559297; /**********************************************************************/ /* Rand */ /*--------------------------------------------------------------------*/ /* Minimal standard random number generator as defined in: */ /* "Random Number Generators: Good ones are hard to find", */ /* Communications of the ACM, 31(1988)1192. */ /**********************************************************************/ #define A 16807 #define M 2147483647 #define Q 127773 #define R 2836 double Rand(void) { int hi, lo, test; hi = Seed / Q; lo = Seed % Q; test = A * lo - R * hi; if (test > 0) Seed = test; else Seed = test + M; return((double)Seed / (double)M); } **************************************************************** From: dario@rs5.csrsrc.mi.cnr.it (Dario Bressanini) Dear Dave, If I may add something to what Brian Hammond wrote, if you are going to do Quantum Monte Carlo for Electronic structure, the time you spend computing the value of the trial wavefunction is much more than the time you spend to generate random deviates I am working on QMC since 1991 and the only thing i can say is: don't use a system provided generator: they usually have a too short cycle. Happy Quantum Monte Carlo computing. dario bressanini dario@rs5.csrsrc.mi.cnr.it **************************************************************** From: Kari Kankaala Dear Dave, Your posting was forwarded to me by a colleague. Our group has been recently playing much with pseudorandomnumber generation, testing a variety of generators etc (e.g. Phys. Rev. Lett, 73, 2513 (1994), Phys. Rev. E, 48, R4211 (1993), see also Nature, 372, Dec 1, 1994, p. 403). I'll try to comment your questions briefly below. > > I am starting to do Quantum Monte Carlo calculations and >I am interested in the performance of random number generators. Our interest in the work on PRNGs was due to the fact that we mainly weork on classical MC simulations of Stat Mech Systems (surfaces, polymers, complex systems in general). > > There have been a number of tests developed, the most trusted >seem to be those in "Seminumerical Algorithms" by Knuth. And there >are a number of supplemental random number generators, such as the >ran3 generator in "Numerical Recipies". It has also been demonstrated >that in the past the random number generators built into various >programming languages were not to be trusted. > Do not touch RAN3. It is one of the worst we tested. > My question is - Do the random number generators built into >the current generation of compilers pass statistical tests? > Generally, do not use any built in generators unless you have demonstrated data that these generators are "good". Cray has a decent generator in Fortran library but e.g. Convex has a extremely poor one available. > I have tested the built in random number generators in >Borland C++ and GNU C++ and so far they have passed all tests >(working out of Knuth). I would like to know if it is only C++ that >has a reliable random number generator implemented or if all >compilers are now reliable, or if it is only certain vendors even within >the C++ language? The language does not matter. It is a question of converting the algorithm into a code correctly. It can be only correct or incorrect, nothing in between. This can be done in any language. If you have disrecpancies with two implementations of the same algortihm, it would suggest that one of them is wrong. Whether it is the C++,C, Fortran or wahtever version, has to be checked. What comes to vendor supplied PRNGs, I would reiterate the earlier comment, that never use those unless someone has demonstarted to you, how reliable both the algorithm and the vendor implementation of them are. > > The reason that I would like to use built in random number >generators is that Knuth indicates that code written directly in >assembly language can accomplish the task faster than code written in >standard source code. The GNU C++ built in random number generator >is a factor of 2 faster than the ran3 generator using all available >compiler optimization. > You can find good algorithms and references to them e.g. in our papers. I can provide you with some codes if you need them. But basically relying on vendor supplied stuff, is like relying on a variety of black boxes and then claiming that your results are significant even if you do not know all the steps in between. Hope this scared you from using vendor supplied stuff :-) But putting pressure on vendors, would be a very good thing. Still, most vendors consider random number generation as a some kind of sidetrack, and do not understand its meaning in modern modelling and simulation. Questions? Comments? I'd be happy to reply. Regards, -kari kankaala **************************************************************** Thank you to everyone who replied. I hope this helps some others as much as it has helped me. Dave Young young@slater.cem.msu.edu youngdc@msucem -------------------------------------------------------------------------- It can be easily shown that certain methods are not applicable to solving the many particle Schrodinger equation. However, it has not been shown that it is impossible to find an analytic solution. The only question remaining is "Who will have the honor?" -------------------------------------------------------------------------- -------This is added Automatically by the Software-------- -- Original Sender Envelope Address: young@slater.cem.msu.edu -- Original Sender From: Address: young@slater.cem.msu.edu CHEMISTRY@ccl.net -- everyone | CHEMISTRY-REQUEST@ccl.net -- coordinator MAILSERV@ccl.net: HELP CHEMISTRY | Gopher: www.ccl.net 73 Anon. ftp www.ccl.net | CHEMISTRY-SEARCH@ccl.net -- archive search http://www.ccl.net/chemistry.html | for info send: HELP SEARCH to MAILSERV From chemistry-request@server.ccl.net Wed Feb 23 06:36:37 2000 Received: from iitkgp.iitkgp.ernet.in ([202.141.127.2]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id GAA12288 for ; Wed, 23 Feb 2000 06:36:31 -0500 Received: from cts.iitkgp.ernet.in (cts.iitkgp.ernet.in [144.16.192.220]) by iitkgp.iitkgp.ernet.in (8.9.3/8.9.3) with ESMTP id QAA12546 for ; Wed, 23 Feb 2000 16:00:29 -0500 (GMT) Received: from localhost (rajarshi@localhost) by cts.iitkgp.ernet.in (8.9.3/8.9.3) with ESMTP id PAA15155 for ; Wed, 23 Feb 2000 15:53:54 +0530 X-Authentication-Warning: cts.iitkgp.ernet.in: rajarshi owned process doing -bs Date: Wed, 23 Feb 2000 15:53:54 +0530 (IST) From: Rajarshi Guha X-Sender: rajarshi@cts.iitkgp.ernet.in To: CompuChem List Subject: FAQ? Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi, I've noticed that a lot of meesages contain the similar questions.I don't mean to put down the people who are asking these questions but would'nt it be better for there to be a FAQ or something which would contain ready made answers to many simple/standard questions (especially the ones about specific algorithms,text references and software websites etc).Then a simple search through the FAQ would provide the answers. Thsi would entail that somebody would have to take the responsibility of looking after this FAQ.(Personally I would offer to do so - but I don't have the requird knowledge base to really make a full and in depth FAQ.If somebody needs help I'd be willing assist -- this assumes that such a FAQ does not exist right now!) Bye, =============================================================== Rajarshi Guha | ...but there was no one Dept. of Chemistry | in it........ IIT Kharagpur. | | RG. email: rajarshi@presidency.com | =============================================================== From chemistry-request@server.ccl.net Wed Feb 23 07:36:50 2000 Received: from hydra.chem.rug.nl (hydra.chem.rug.nl [129.125.35.229]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id HAA13495 for ; Wed, 23 Feb 2000 07:36:49 -0500 Received: from [129.125.7.13] (theochem1.chem.rug.nl [129.125.7.13]) by hydra.chem.rug.nl (AIX4.2/UCB 8.7/8.7) with ESMTP id MAA31498 for ; Wed, 23 Feb 2000 12:29:44 +0100 (MET) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Sender: swart@hydra.chem.rug.nl Message-Id: In-Reply-To: Date: Wed, 23 Feb 2000 12:33:49 +0100 To: CHEMISTRY@ccl.net From: Marcel Swart Subject: Re: CCL:FAQ? >Hi, > I've noticed that a lot of meesages contain the similar questions.I >don't mean to put down the people who are asking these questions but >would'nt it be better for there to be a FAQ or something which would >contain ready made answers to many simple/standard questions (especially >the ones about specific algorithms,text references and software websites >etc).Then a simple search through the FAQ would provide the answers. > >Thsi would entail that somebody would have to take the responsibility of >looking after this FAQ.(Personally I would offer to do so - but I don't >have the requird knowledge base to really make a full and in depth FAQ.If >somebody needs help I'd be willing assist -- this assumes that such a FAQ >does not exist right now!) Isn't it true, that one can search the CCL archives ? If so, no need to reinvent the wheel. Marcel. ============================================================== Marcel Swart E-mail : m.swart@chem.rug.nl ============================================================== From chemistry-request@server.ccl.net Wed Feb 23 09:05:39 2000 Received: from theory.tc.cornell.edu (THEORY.TC.CORNELL.EDU [128.84.30.174]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id JAA13949 for ; Wed, 23 Feb 2000 09:05:39 -0500 Received: from tc.cornell.edu (LIGAND.TC.CORNELL.EDU [128.84.20.182]) by theory.tc.cornell.edu (8.8.4/8.8.3/CTC-1.0) with ESMTP id HAA22403 for ; Wed, 23 Feb 2000 07:58:33 -0500 Sender: richard@tc.cornell.edu Message-ID: <38B3D874.B379B053@tc.cornell.edu> Date: Wed, 23 Feb 2000 07:54:12 -0500 From: Richard Gillilan X-Mailer: Mozilla 4.05C-SGI [en] (X11; I; IRIX64 6.5 IP30) MIME-Version: 1.0 To: Chemistry@CCL.net Subject: Re: CCL:The "birth-year" of computational chemistry? References: <38B38C76.C0CE0DC8@mdli.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit > I have two questions: > > 1. Who published the first paper about using an electronic > computer to solve a(ny) chemical problem? I seem to recall, that Monte Carlo methods invented during the Atom Bomb project in the 40's may have been the first electronic computer calculations on condensed matter, although they may not qualify as chemistry. I remember a facinating talk by W. Wood (Los Alamos) at a computer simulation conference at University of Utah about 10 years ago entitled "The Early History of Computer Simulations". I'll try to dig up his paper. Richard Gillilan Cornell Theory Center From chemistry-request@server.ccl.net Wed Feb 23 09:34:29 2000 Received: from beta.chem.msu.ru (beta.chem.msu.ru [195.208.208.11]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id JAA14111 for ; Wed, 23 Feb 2000 09:34:23 -0500 Received: from analyt4711 (analyt471-1.chem.msu.ru [195.208.209.168]) by beta.chem.msu.ru (8.9.3/8.9.3) with SMTP id QAA27222; Wed, 23 Feb 2000 16:26:58 +0300 (MSK) Message-ID: <00b201bf7e01$cf9b45f0$a8d1d0c3@chem.msu.ru> From: "Michael G. Razumov" To: Cc: , , Subject: SUMMARY: internal coordinates to cartesian Date: Wed, 23 Feb 2000 16:27:51 +0300 X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2014.211 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2014.211 Hi! Several days ago I asked a question about a program, which transferes internal molecule coordinates to cartesian. Thanks to E.Hiramatsu; David Enot and Crowther, Robert L {DISC~Nutley} for the help. There are two programs to do the subj: BABEL can transfere MOPAC internal coordinates file to cartesian XYZ file. See http://www.eyesopen.com/babel.html You can also use Molden to translate your Z matrix to cartesian coordinates. See http://www.caos.kun.nl/~schaft/molden/molden.html You can find the internal coordinates format (Z matrix) for example at: http://www.chm.tu-dresden.de/edv/mopac6/mop.html Bye. ---------------------------------------------------------------------------- --------- Michael G. Razumov, Postgraduate Student of Kurnakov Institute of General and Inorganic Chemistry and Lomonosov Moscow State University E-Mail: michraz@analyt.chem.msu.ru michael@analyt.chem.msu.su ICQ UIN: 25169010 From chemistry-request@server.ccl.net Wed Feb 23 11:07:15 2000 Received: from smtp.pace.edu (ntutil.pace.edu [205.232.111.9]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id LAA14776 for ; Wed, 23 Feb 2000 11:07:14 -0500 Received: from fsmail.pace.edu (205.232.111.6:1709) by smtp.pace.edu (LSMTP for Windows NT v1.1b) with SMTP id <0.A7DF2866@smtp.pace.edu>; Wed, 23 Feb 2000 9:57:52 -0500 Date: Wed, 23 Feb 2000 09:57:52 -0500 Message-Id: <200002230957.AA592183852@fsmail.pace.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii From: "Frank Marchese" Reply-To: To: Subject: Re: CCL:The "birth-year" of computational chemistry? X-Mailer: What do you mean by "electronic" ? Are the computer's analog or digital? Lord Kelvin invented/ used an analog computer called the "differential analyzer" to solve 2nd order differntial equations, in particular for analysis of tides. In the late 1920s, Vannevar Bush of MIT re-invented Kelvin's analog machine on a grander scale. Douglas Hartree spent time with Bush, took the basic plans back to England and built his own. During the 1930s, Bush's, Hartree's, and similar machines built in physics departments were used to solve the Schrodinger equation. The best history on this stuff remains Herman Goldstone's book: The Computer from Pascal to von Neumann (Princeton U. Press) See the references within. Francis T. Marchese Professor Dept. of Computer Science Pace University ---------- Original Message ---------------------------------- From: Lingran Chen Date: Tue, 22 Feb 2000 23:29:58 -0800 >Dear CCLers: > >In the Encyclopedia of Computational Chemistry (ECC), there are some >interesting >statements about the "birth-year" of computational chemistry. For >example, > >1. "Computational chemistry has existed for half a centrury, ..." > - Forword of ECC by Prof. Pople (1997). > >Thus, the birth-year of CC is about 1947. > >2. "Although its origin date back much earlier, > Computational Chemistry evolved into a distinctly discernible >discipline > three decades ago ..." > - Preface of ECC by Prof. Schleyer (1997?) > >Thus, the birth-year of CC is about 1967. > >3. "The general availability of computers at major research laboratories >and > the availability of FORTRAN as a general programming language to make >use of > these computers marks > 1960 as approximately the beginning of modern computational chemistry >>..." > - "HISTORY OF COMPUTATIONAL CHEMISTRY: A PERSONAL VIEW" > by prof. Streitwieser. (ECC, p.1238) > >Here, the birth-year of CC is 1960. > >I have two questions: > >1. Who published the first paper about using an electronic >computer to solve a(ny) chemical problem? > >In my personal record, the earlest publication is: > >P D. Zemany, "Punched Card Catalog of Mass Spectra Useful in Qualitative >Analysis," >Anal. Chem., 22, 1950, 920. > >2. Who wrote the first quantum chemistry program? > >The answers to the above questions may indicate the "reasonable" >birth-year of CC. >Thanks in advance. > >-Lingran > >*********************************** >Lingran Chen, Ph.D. >Advisory Scientist >MDL Information Systems, Inc. >14600 Catalina Street >San Leandro >CA 94577 >U.S.A. > >Phone: (510) 895-1313 >FAX: (510) 614-3616 > >Email: LCHEN@MDLI.COM >Web: http://www.mdli.com >*********************************** > >-= This is automatically added to each message by mailing script =- >CHEMISTRY@ccl.net -- To Everybody | CHEMISTRY-REQUEST@ccl.net -- To Admins >MAILSERV@ccl.net -- HELP CHEMISTRY or HELP SEARCH >CHEMISTRY-SEARCH@ccl.net -- archive search | Gopher: gopher.ccl.net 70 >Ftp: ftp.ccl.net | WWW: http://www.ccl.net/chemistry/ | Jan: jkl@ccl.net > > > > > > From chemistry-request@server.ccl.net Wed Feb 23 11:08:43 2000 Received: from schrodinger.com (root@thermidore.schrodinger.com [192.156.98.99]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id LAA14789 for ; Wed, 23 Feb 2000 11:08:42 -0500 Received: from sally.schrodinger.com (shenkin@sally.schrodinger.com [166.84.149.25]) by schrodinger.com (8.8.5/8.8.5) with ESMTP id HAA18564 for ; Wed, 23 Feb 2000 07:01:37 -0800 Received: from localhost (shenkin@localhost) by sally.schrodinger.com (8.8.8/8.8.5) with ESMTP id KAA07861 for ; Wed, 23 Feb 2000 10:01:37 -0500 X-Authentication-Warning: sally.schrodinger.com: shenkin owned process doing -bs Date: Wed, 23 Feb 2000 10:01:36 -0500 (EST) From: Peter Shenkin To: chemistry@ccl.net Subject: Re: CCL:FAQ? In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi, Here's a proposed FAQ: Q: I have a question about how to use Gaussian. A: Ask Gaussian. :-) (yes, just joking), -P. -- ** Whether the playing field is level depends on the coordinate system. *** ********* Peter S. Shenkin; Schrodinger, Inc.; (201)433-2014 x111 ********* *********** shenkin@schrodinger.com; http://www.schrodinger.com *********** From chemistry-request@server.ccl.net Wed Feb 23 09:15:06 2000 Received: from yogi.pc1.uni-duesseldorf.de (yogi.pc1.uni-duesseldorf.de [134.99.152.44]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id JAA14017 for ; Wed, 23 Feb 2000 09:15:06 -0500 Received: (from jochen@localhost) by yogi.pc1.uni-duesseldorf.de (8.9.3/8.9.3) id OAA07559; Wed, 23 Feb 2000 14:08:32 +0100 From: "Jochen Küpper" MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Message-ID: <14515.56269.328317.689516@yogi.pc1.uni-duesseldorf.de> Date: Wed, 23 Feb 2000 14:08:29 +0100 (CET) To: Rajarshi Guha Cc: CompuChem List Subject: [FAQ] CCL:FAQ? In-Reply-To: References: X-Mailer: VM 6.72 under 21.1 (patch 4) "Arches" XEmacs Lucid Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by server.ccl.net id JAA14018 Rajarshi Guha writes: > Thsi would entail that somebody would have to take the responsibility of > looking after this FAQ.(Personally I would offer to do so - but I don't > have the requird knowledge base to really make a full and in depth FAQ.If > somebody needs help I'd be willing assist -- this assumes that such a FAQ > does not exist right now!) Well, go ahead and start the FAQ. Put it up on CCL, post it here once a month, and ask people to contribute questions with answers they want to see in there :-) Probably you start with: Q: Where can I find code to solve my numerical problem XYZ ? A: Check out http://www.netlib.org. Another nice interface is at http://gams.nist.gov. Jochen -- Heinrich-Heine-Universität Institut für Physikalische Chemie I Jochen Küpper Universitätsstr. 1, Geb. 26.43.02.29 40225 Düsseldorf, Germany phone ++49-211-8113681, fax ++49-211-8115195 http://www.jochen-kuepper.de From chemistry-request@server.ccl.net Wed Feb 23 11:05:20 2000 Received: from ns02.sbphrd.com (firewall-user@ns02.sbphrd.com [208.198.64.2]) by server.ccl.net (8.8.7/8.8.7) with SMTP id LAA14764 for ; Wed, 23 Feb 2000 11:05:19 -0500 Received: by ns02.sbphrd.com; id JAA17427; Wed, 23 Feb 2000 09:58:12 -0500 Received: from unknown(139.136.64.5) by ns02.sbphrd.com via smap (V5.0) id xma017405; Wed, 23 Feb 00 09:58:04 -0500 Received: from phufs1.um.us.sbphrd.com (phufs1.um.us.sbphrd.com [139.136.98.91]) by phinet.sbphrd.com (8.9.1b+Sun/8.9.1) with ESMTP id JAA07303 for ; Wed, 23 Feb 2000 09:58:20 -0500 (EST) Received: from phu847.um.us.sbphrd.com (phu847.um.us.sbphrd.com [139.136.98.147]) by phufs1.um.us.sbphrd.com (8.8.8/8.8.8/sbphrd/mh/19990803) with ESMTP id JAA11616 for ; Wed, 23 Feb 2000 09:58:03 -0500 (EST) From: M Dominic Ryan Received: (from ryanmd@localhost) by phu847.um.us.sbphrd.com (8.8.8/8.8.8) id JAA51131 for Chemistry@CCL.net; Wed, 23 Feb 2000 09:58:03 -0500 (EST) Date: Wed, 23 Feb 2000 09:58:03 -0500 (EST) Message-Id: <200002231458.JAA51131@phu847.um.us.sbphrd.com> To: Chemistry@CCL.net Subject: The "birth-year" of computational chemistry? I dug into this question about 12 years ago. It depends of course on just what you call computational chemistry. Does it have to be using a computer? I would argue that should not be a requirement. Rather the real underlying field is the construction of a numeric simulation that represents 'reality' in a context that permits testing concepts and thereby learning something. Today we use computers in their current form, perhaps this will change in the decades to come. I point back to early spectroscopic work. The earliest I think is work by Andrews, in the 20s. Once citation I have is in Physical Reviews, v36, p531, 1930, Kettering, Shutts and Andrews (General Motors Corporation). "A representation of the Dynamic Properties of Molecules by Mechanical Models". I think of this as the birth of molecular modeling. They built molecular models of things like toluene, benzene, chlorobenzene using steel balls and calibrated springs. They attached an oscillator and swept the frequency. They then used a strobe light an found the standing waves. In this way they mapped out a molecular mechanics model the IR and Raman spectra. They were then able to draw conclusions about the correct molecular structure by doing the experiment on various bonding models of benzene by having springs attached in different arrangements. I think they got it wrong in the end, but that sure looks like molecular mechanics with an analog computer to me. There is also the work by Urey and Bradley, whose names are still attached to a force field, that begins around 1931 with a paper in Physical Review, p1969, 1931 (apparently Bradley's Ph.D. thesis). He discusses Andrew's papers and also a paper I never saw by Bjerrum in physik. Ges. Ber. v16, p737, 1914, proposing those sorts of forces acting upon bonded atoms. He of course also cites Slater (1931) and Pauling JACS,v53,p1367, 1931, as discussing the nature of the forces on a chemical bond. Still later, I put Frank Westheimer (with John Mayer) at the birth of computational chemistry using numerical methods. He calculated an activation energy for racemization of 2,2'-dibromo-4,4'-dicarboxydiphenyl in J. Chemical Physics, v15, p252, 1947 (there is also an earlier paper v14, p733,1946). He used an exponential function to describe the rates and pull out constants. Other names at that time are T.L. Hill, J. Chem. Phys. v14,465 (1946) and W. Gordy, J.Chem.Phys 14,305(1946). __ M. Dominic Ryan (610)-270-6529 SmithKline Beecham Pharmaceuticals Internet: ryanmd@mh.us.sbphrd.com King of Prussia, PA From chemistry-request@server.ccl.net Wed Feb 23 12:14:51 2000 Received: from mailrelay2.lrz-muenchen.de (mailrelay2.lrz-muenchen.de [129.187.254.102]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id MAA15509 for ; Wed, 23 Feb 2000 12:14:50 -0500 Received: from olymp.cup.uni-muenchen.de by mailrelay2.lrz-muenchen.de; Wed, 23 Feb 2000 17:07:42 +0100 Received: by olymp.cup.uni-muenchen.de (AIX 3.2/UCB 5.64/4.03) id AA59981; Wed, 23 Feb 2000 17:07:38 +0100 Date: Wed, 23 Feb 2000 17:07:37 +0100 (MET) From: Stefan Seidler To: jmmckel@attglobal.net Cc: chemistry@ccl.net Subject: Error Function [ERF] Code In-Reply-To: <38B256A8.9AA0BCCF@attglobal.net> Message-Id: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On Tue, 22 Feb 2000 jmmckel@attglobal.net wrote: > their own way of computing this information. Can someone please point > me to a source for a fortran code to do compute ERF, however slow it > might be? > Look at Chapter 6.2 in Numerical Recipes, available online: http://www.ulib.org/webRoot/Books/Numerical_Recipes/bookfpdf/f6-2.pdf Regards, Stefan Stefan.Seidler@cup.uni-muenchen.de From chemistry-request@server.ccl.net Wed Feb 23 12:58:12 2000 Received: from bacchus.pc1.uni-duesseldorf.de (root@bacchus.pc1.uni-duesseldorf.de [134.99.152.11]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id MAA15899 for ; Wed, 23 Feb 2000 12:58:11 -0500 Received: (from jochen@localhost) by bacchus.pc1.uni-duesseldorf.de (8.8.7/8.8.7) id RAA17505; Wed, 23 Feb 2000 17:50:43 +0100 From: "Jochen Küpper" MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Message-ID: <14516.4067.58321.523141@bacchus.pc1.uni-duesseldorf.de> Date: Wed, 23 Feb 2000 17:50:43 +0100 (CET) To: Peter Shenkin Cc: chemistry@ccl.net Subject: [FAQ] CCL:FAQ? In-Reply-To: References: X-Mailer: VM 6.72 under 21.1 (patch 8) "Bryce Canyon" XEmacs Lucid Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by server.ccl.net id MAA15900 > Here's a proposed FAQ: > > Q: I have a question about how to use Gaussian. > A: Ask Gaussian. > > :-) (yes, just joking), Well, no joke: Q: I have a question about how to use Gaussian. A: a) Check the Gaussian manual. b) Look up the information at http://www.gaussian.com. c) Check Q&A "20-98". (reference to more specific Q&A about typical questions concerning Gaussian within the FAQ.) :-) Jochen -- Heinrich-Heine-Universität Institut für Physikalische Chemie I Jochen Küpper Universitätsstr. 1, Geb. 26.43.02.29 40225 Düsseldorf, Germany phone ++49-211-8113681, fax ++49-211-8115195 http://www.jochen-kuepper.de From chemistry-request@server.ccl.net Wed Feb 23 15:20:30 2000 Received: from ccl.net (atlantis.ccl.net [192.148.249.4]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id PAA16681 for ; Wed, 23 Feb 2000 15:20:30 -0500 Received: from krakow.ccl.net (krakow.ccl.net [192.148.249.195]) by ccl.net (8.8.6/8.8.6/OSC 1.1) with ESMTP id OAA15085; Wed, 23 Feb 2000 14:13:17 -0500 (EST) Date: Wed, 23 Feb 2000 14:13:18 -0500 (EST) From: Jan Labanowski To: chemistry@ccl.net cc: Jan Labanowski Subject: FAQ and Search on CCL Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Dear, As to CCL FAQ and searches... FAQ... I would welcome it as a community project... My problem is with time and the staff. The students who help me with CCL are computer science students, not the Computational Chemistry oriented. So I cannot expect them to help me with Computational Chemistry. But you can... When You see a FAQ question, and you respond to it in more or less "closed" form (i.e., not with a one liner, but something which will help a person who is not a guru), please copy it to me (jkl@ccl.net) with a note "Add to FAQ" or upload it to incoming of our anonymous ftp site: http://www.ccl.net/chemistry/aboutccl/contributing/ After I have some critical mass of these FAQ topics/answers, I will put it on the web. I may even create a web form for FAQ submissions later. Please help. I am short handed so do not expect me to act as an editor... Just put the stuff in a closed form, so I can pass it to a student who does not know what it is, and he/she can edit it into the FAQ tree. I really think that this would improve the quality of the list and discussions in general. It would be best, if you could make these contributions in HTML. All the links sould be also displayed, e.g.: http://www.ccl.net/chemistry The format can be as: Author: your name and e-mail address Date: the current date Topic: Keywords relevant to the topic below Question: The precise question to which you provide answer Answer: The gory details, but less than a few pages. If answer needs more then few pages of narrative, it is not an answer to a Question, it is a review (which I always welcome and you have some o them under "Document Archives"). You can of course have some examples or accompanying files for the FAQ, please attach them. The other option is, make it a tar or zip, and upload it to the incoming directory on ftp.ccl.net and let me know. If you have a better idea, send it to me, and I will choose the thing which best fits the overall structure of CCL. Do not expect answers soon, since I will be at Sanibel for a week. SEARCHES.... Sorry... Our htdig search started to produce indices which want to be larger than 2Gigs of space. Of course, a no, no under older Linices which does not like files larger than 2Gigs (yes, we are upgrading slowly). So, we have to find the solution. In the meantime use the Regular Expression search. Sorry, it is slow (and often dies), but this is what we have at this moment. My students are looking at alternatives/solutions. Thank you for your patience... Jan jkl@ccl.net Jan K. Labanowski | phone: 614-292-9279, FAX: 614-292-7168 Ohio Supercomputer Center | Internet: jkl@ccl.net 1224 Kinnear Rd, | http://www.ccl.net/chemistry.html Columbus, OH 43212-1163 | http://www.ccl.net/ From chemistry-request@server.ccl.net Wed Feb 23 15:23:50 2000 Received: from ivory.trentu.ca (trentu.ca [192.75.12.103]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id PAA16761 for ; Wed, 23 Feb 2000 15:23:49 -0500 Received: from trentu.ca ([204.225.13.50]) by trentu.ca (PMDF V5.2-32 #29543) with ESMTP id <01JM9099L930000SL5@trentu.ca> for chemistry@ccl.net; Wed, 23 Feb 2000 14:16:30 EST Date: Wed, 23 Feb 2000 14:20:15 -0500 From: elewars Subject: PPP METHOD SUMMARY To: chemistry@ccl.net Message-id: <38B432EE.9FAD8915@trentu.ca> MIME-version: 1.0 X-Mailer: Mozilla 4.7 [en] (WinNT; I) Content-type: text/plain; charset=us-ascii Content-transfer-encoding: 7bit X-Accept-Language: en Wed 2000 Feb 23 SUMMARY OF STATUS OF PPP METHOD Hello, here are the answers to my question about the status of the Pariser-Parr-Pople method of doing calculations on pi systems. Thanks very much to all who answered. EL ------------- The Question: Subject: CCL:PPP Method: status? Date: Wed, 16 Feb 2000 19:04:11 -0500 From: elewars 2000 Feb 16 Subject: the Pariser-Parr-Pople method Could someone with some familiarity with the PPP method please tell me: 1) is it still used at all? 2) was it used for calculation of UV spectra *at a given geometry* only, or could it perform realistic geometry optimizations on pi-systems? 3) what has it evolved into: what currently popular method or program for calculating electronic spectra is an offspring of the PPP method? Thank you. E. Lewars ============= The Answers (#1--#6) #1 Wed, 16 Feb 2000 21:06:31 +0000 From: jmmckel@attglobal.net PPP theory is still used. It can be quite accurate, and very fast, when parameterised carefully. It is best used for planar molecules in general. If nonplanar configurations are considered one must take care to avoid twists that bring components of localised lone pairs out of the plane. It works best at a frozen geometry. It can be used to optimize bond lengths by calibrating the expression R = a + b*P where P is the computed bond order for each nearest-neighbor bond type at experimental or high quality ab initio geometries. PPP theory is just the pi part of INDO/S and similar methodologies, and assumes a frozen sigma field that is incorporated into the one center parameters. Because INDO/S and related all valence electron methodologies treat sigma electrons explicitly it is superior to PPP theory for pnon-planar molecules. Geometry optimization for INDO/S spectroscopic methods are not very well developed because parameterizations for spectra give poor geometries, in general. John McKelvey ===== #2 Date: Wed, 16 Feb 2000 19:58:09 -0500 From: "David A. Pensak" "David A. Pensak" The best source of information about this is none other than Rudy Pariser. He is rudypar@aol.com. He lives here in Delaware having retired from DuPont a few years ago. I'm sure he would be most happy to answer your questions. ========= #3 Date: Thu, 17 Feb 2000 09:49:15 +0100 From: Stefan Grimme Hello, I have written a PPP program which does PPP/SCI and PPP/RPA calculations for UV and CD spectra. I'm using it for large supramolecular structures. It automatically discards all non-pi atoms from a cartesian input file for the whole molecule and assigns the parameters. Non-planarity effects and through-space interactions are treated explicitely. Its a development version but you can have it if you want. Stefan Grimme ___________________________________________ Prof. Dr. Stefan Grimme Organisch-Chemisches Institut (Abt. Theoretische Chemie) Westfaelische Wilhelms-Universitaet Corrensstrasse 40 D-48149 Muenster Tel (+49)-251-83 36512/33241 Fax (+49)-251-83 36515 Email:grimmes@uni-muenster.de ___________________________________________ ========= #4 Date: Thu, 17 Feb 2000 11:15:49 +0000 From: "Charlton, Michael" Pi-System from Rudolf Naef performs PPP calcs to predict UV/Vis spectra. Michael Charlton. ========== #5 Date: Thu, 17 Feb 2000 07:51:27 -0500 From: "Risser, Steven M" Hi, As far as PPP goes, it is still used by several groups doing polymer physics. The primary solution method now involves full-CI like calculations, looking at optical excitations. PPP is NOT useful for geometries, since you are completely neglecting the sigma bonded electrons. There had been work 10-15 years ago on representing the sigma bonds by harmonic oscillators, and looking at distortions due to the p bonding, but that was not too successful. As far as available programs, I do not know of any that I would say is a direct descendant of PPP. Most of the semi-empirical methods come from the NDDO school, which is somewhat distinct. The other related work is in physics, where a lot of electronic structure is done using the Hubbard model (only nearest-neighbor electronic interactions). Hope this helps =============== #6 Date: Thu, 17 Feb 2000 14:25:30 +0100 From: Jens Spanget-Larsen Organization: Roskilde Universitetscenter E. Lewars: > Subject: the Pariser-Parr-Pople method > Could someone with some familiarity with the PPP method please tell me: > 1) is it still used at all? Until quite recently, highly parametrized PPP models were used in the dye industry, see for instance the publications by J. Griffiths from the Dept. of Colour Chemistry and Dyeing at Leeds University (UK): - "Practical aspects of colour prediction of organic dye molecules", Dyes and Pigments, 1982, 3:211-233 - "Modern Dye Chemistry", Chemistry in Britain, 1986, 997-1000 - "Funktionelle Farbstoffe", Chemie in unserer Zeit, 1993, 12:21-31 > 2) was it used for calculation of UV spectra *at a given geometry* only, > or could it perform realistic geometry optimizations on pi-systems? The classic PPP pi-electron method is a combination of Pariser and Parr's CI model and Pople's SCF-MO model and was designed for prediction of pi-pi* electronic transitions, assuming a 'given geometry'. In order to 'perform realistic geometry optimizations' you obviously need to consider explicitly the role of the sigma-core. Several hybrid models, describing the sigma-system by a mechanical force-field model and the pi-system by a Pople-like SCF-MO method, have been proposed for the prediction of geometries and heats of formation for conjugated systems. For an account of M.J.S. Dewar's work in this field, see his book: "The Molecular Orbital Theory of Organic Chemistry", Chapter 5 ('The Pople Method'). For modern implementations, see for example the treatise by J.J. Gajewski, K.E. Gilbert, and J. McKelvey in "Advances in Molecular Modeling", 1990, 2:65-92, and lit. cited therein. > 3) what has it evolved into: what currently popular method or program > for calculating electronic spectra is an offspring of the PPP method? J.A. Pople extended his pi-electron SCF-procedure to descriptions involving all valence electrons, resulting, of course, in the all-valence NDO procedures CNDO, INDO, etc. Corresponding procedures CNDO/S and INDO/S were developed and parametrized for the prediction of electronic transitions. Today, these procedures have essentially replaced the classic PPP method. The most popular model in this field is probably ZINDO/S, developed by the late M.C. Zerner and his group in Florida; ZINDO/S is included, for example, in the current versions of HyperChem and Gaussian. Yours, Jens >--< PS. One beauty of the standard PPP method is the presence in this model of perfect orbital and state 'pairing symmetry' for alternant hydrocarbons. All-valence NDO procedures like CNDO/S and INDO/S break down this symmetry; however, the breakdown tends to be too strong, sometimes with unfortunate consequences for the predicted transitions. See Spanget-Larsen: Theor. Chem. Acc. 1997, 98:137-153. =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= JENS SPANGET-LARSEN Phone: +45 4674 2000 (RUC) Department of Chemistry +45 4674 2710 (direct) Roskilde University (RUC) Fax: +45 4674 3011 P.O.Box 260 E-Mail: JSL@virgil.ruc.dk DK-4000 Roskilde, Denmark http://www.rub.ruc.dk/dis/chem/psos/ =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= =========== ========== From chemistry-request@server.ccl.net Tue Feb 22 18:22:15 2000 Received: from hotmail.com (f94.law7.hotmail.com [216.33.237.94]) by server.ccl.net (8.8.7/8.8.7) with SMTP id SAA08840 for ; Tue, 22 Feb 2000 18:22:14 -0500 Received: (qmail 42783 invoked by uid 0); 22 Feb 2000 22:13:18 -0000 Message-ID: <20000222221318.42782.qmail@hotmail.com> Received: from 164.138.135.194 by www.hotmail.com with HTTP; Tue, 22 Feb 2000 14:13:18 PST X-Originating-IP: [164.138.135.194] From: "Alexandre Hocquet" To: CHEMISTRY@ccl.net Subject: answer to NBO intermolecular interactions Date: Tue, 22 Feb 2000 22:13:18 GMT Mime-Version: 1.0 Content-Type: text/plain; format=flowed Thank you very much to the community and to F. Weinhold himself for correcting the erroneous reference : a.. Intermolecular Interactions From A Natural Bond Orbital, Donor-Acceptor Viewpoint. A. E. Reed, L. A. Curtiss, and F. Weinhold, Chem. Revs. 88, 899 (1988) and not 1998 ...time flies... ------------------------------------------------------------ Alexandre HOCQUET Laboratoire de Physicochimie Biomoléculaire et Cellulaire ESA CNRS 7033 hocquet@lpbc.jussieu.fr Fax: 0144277560 (do not dial first 0 from abroad) LPBC, case courrier 138 4 Place Jussieu, 75252 PARIS Cedex 05 France ------------------------------------------------------------ ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From chemistry-request@server.ccl.net Tue Feb 22 19:51:18 2000 Received: from hotmail.com (f205.law3.hotmail.com [209.185.241.205]) by server.ccl.net (8.8.7/8.8.7) with SMTP id TAA09382 for ; Tue, 22 Feb 2000 19:51:17 -0500 Received: (qmail 32642 invoked by uid 0); 22 Feb 2000 23:42:20 -0000 Message-ID: <20000222234220.32641.qmail@hotmail.com> Received: from 128.101.162.4 by www.hotmail.com with HTTP; Tue, 22 Feb 2000 15:42:20 PST X-Originating-IP: [128.101.162.4] From: "Y Z" To: CHEMISTRY@ccl.net Cc: dkwmok@fg702-6.abct.polyu.edu.hk Subject: Summary of 3D-FFT Date: Tue, 22 Feb 2000 18:42:20 EST Mime-Version: 1.0 Content-Type: text/plain; format=flowed After I asked the question for 3D-FFT, I got a lot of responses. Thanks a lot for their helps , my problem have been solved. Here is the summary My original question is: Hi, Are there any way to calculate a (256,256,256) order DOUBLE COMPLEX FFT in FORTRAN77? I tried to do the following subrouitine but cannot pass the compiling. By f77 in SGI : Warning: Stack frame size (270015328) larger than system limit (67108864) +++++++++++++++++++++++++++++++++ program test IMPLICIT DOUBLE PRECISION (A-H,O-Z) DOUBLE COMPLEX AA(256,256,256) DIMENSION ID(64),CC(64) DIMENSION NORD(3) DO 100 i=1,256 DO 100 j=1,256 DO 100 k=1,256 xx=DBLE(I+j+k) AA(i,j,k)=DCMPLX(xx,0.0D0) 100 CONTINUE NORD(1)=8 NORD(2)=8 NORD(3)=8 IERR=0 CALL DFFT3D(dta,NORD,ID,CC,1,IERR) STOP END -------------------------------------- Below is the responses from people: ----------------------------------- From: laidig@pg.com I don't know if you realize this by the array AA is 268435456 bytes long (268 MB). You have exceeded some system limit on how big an area you can have by default. I don't know what machine you are compiling this on, but on some machines and if you are using the c shell you can type limit to get the limits and limit name_of_limit new_value_for_limit. This may work unless your system will not allow you to set it up high enough. Bill P.S. On my SGI Indigo 2 this is what the limit command gives: cputime unlimited filesize unlimited datasize 2097152 kbytes stacksize 524288 kbytes coredumpsize 0 kbytes memoryuse 524288 kbytes vmemoryuse 2097152 kbytes descriptors 2500 threads 1024 From: David Heisterberg Try putting your large array in a common block. It looks like the default allocation is to put your 256MB array on the stack! Naming it in a SAVE statement or compiling with static allocation may work too. Dave Heisterberg From: "Serguei Patchkovskii" > IMPLICIT DOUBLE PRECISION (A-H,O-Z) > DOUBLE COMPLEX AA(256,256,256) This is a rather large array, requiring 256megabytes of memory to store. I presume that you do have enough memory to handle it (and whatever temporary space FFT might need in addition to that). If you do, you'll have to make sure your resource limits are set high enough to allow such array (use limit if you run under csh or descendants; ulimit under sh & friends). It is possible that your system is not configured to handle arrays of this size; in this case you will have to increase the system-wide limits (since you say you are using an SGI, this can be done with systune; see it's manual page). Finally, unless the default settings has been overriden on your computer, SGI compiler will allocate locally declared arrays on stack - so this is the limit you'll be looking for. You can force the compiler to allocate the array in the data segment by either using -static compilation switch, or by placing AA in a common block. /Serge.P From: Iraj Daizadeh check limit (csh/tcsh) on unix box supercede with ulimit stacksize or limit stacksize unlimited iraj. From: David van der Spoel To: Y Z Save Address Subject: Re: CCL:Three dimensional FFT Date: Sun, 20 Feb 2000 22:17:32 +0100 (CET) Use the UNIX limit command to increase your limit. From: jmmckel@attglobal.net Save Address Block Sender Reply-To: jmmckel@attglobal.net To: Y Z Save Address Subject: Re: CCL:Three dimensional FFT Date: Sun, 20 Feb 2000 14:22:27 +0000 I wonder if your system guru couldn't ive some assistance about the stack.. John McKelvey From: Niri Govind there are very good 3d-fft routines as part of the libcomplib.sgimath libraries on the sgis. you can get information on how to use them by looking at the man pages. you will need to link to this library when you compile your program... try man fft, this should give you all the info you need. hope this helps cheers -niri From: Niri Govind hi, there are very good 3d-fft routines as part of the libcomplib.sgimath libraries on the sgis. you can get information on how to use them by looking at the man pages. you will need to link to this library when you compile your program... try man fft, this should give you all the info you need. hope this helps cheers -niri From: "Daniel Mok" I've got a similar message problem in a different fortran subroutine. It seems to be the problem of the allocated memory size. I don't how to increase the mentioned system limit. Please let me know if you got any reply in doing this. Daniel From: "Jochen Küpper" > Are there any way to calculate a (256,256,256) order > DOUBLE COMPLEX FFT in FORTRAN77? I tried to do the > following subrouitine but cannot pass the compiling. Nothing on netlib ? fftw ? Jochen From: Gijs Schaftenaar a double array(256,256,256) cost roughly 134 MB of memory, you probably dont have this. Best Regards, Gijs Schaftenaar From: Sigismondo Boschi DOUBLE COMPLEX AA(256,256,256) The problem is that SGI compiler allocates all automatic variables in the stack that usually is limited and not well swited for large arrays (even if we are in the "program"). The ways to avoid it are: 1. put the array in a common block 2. SAVE AA 3. compile with option "-static" 4. allocate memory dynamically (OK, thats Fortran90...) I would avoid the first if you do not really need it, and I would use the second, except it is just for a short test case - with the third it works just fine. I strongly discourage to enlarge the stack size for such problems. I Hope This Helps! Bye Sigismondo Boschi From: giesen@chem.umn.edu Compile using the -static flag and the error will go away. (not pretty, but it works) Dave From: "Jochen Küpper" > Are there any way to calculate a (256,256,256) order > DOUBLE COMPLEX FFT in FORTRAN77? I tried to do the > following subrouitine but cannot pass the compiling. Nothing on netlib ? fftw ? Jochen From: Huub van Dam Hmm, you are trying to statically allocate an array of 256 MB and your compiler doesn't like it. There are a few options you could try: 1) Find where the system limit is defined and increase it so that it will allow what you want to do. 2) Change the code to run in parallel in a memory distributed fashion so that each node needs to store maximul 64 MB (i.e. your current system limit) 3) Change your code to use disc space in stead of main memory. 4) Think of a smarter approach that avoids doing a 3D FFT at once. 5) Wait for someone else to mail you a ready to go solution. Each option has its pro's and con's and certain requirements: Option 1 is the easiest (no code changes required) but you need to have sufficient (virtual) memory to store an executable with a data space of 256 MB. I.e. you would need to have about 320 MB of (virtual) memory. Note that if you physical memory is smaller than 256 MB your code may run very slowly. And you need to have the root permissions to mess around with the system parameters. For a quick solution though this is probably your best bet. Option 2 requires you to have access to more and one processor at a time. In this case you would need at least 4 processors. Also you need to have access to either a compiler that can generate a parallel executable for you or you need to change the code substantially. In the latter case the global array tools (http://www.emsl.pnl.gov:2080/docs/global/) may be very useful as they allow you to mimic a shared memory parallel system on distributed memory hardware. Also they provide parallelised essential matrix operations. Option 3 require a lot of code changes (and a lot of time). You would have to break up the complete operation in partial operations each of which fits in the memory you have available. The data you don't need in each partial step you store on disk. This may well require a significant rewrite of you algorithm, and will certainly involve a lot of file handling. Option 4 probably requires the most time. It means rethinking your scientific problem and trying to formulate your solution in such a way that you avoid doing the full 3D FFT in one step. Eg. try to write your solution as 3 one dimensional FFT's or something like that. Although this may take the most time, if you expect to do these kinds of calculations often going this way may be best solution. Option 5 requires a significant amount of luck. Huub From: Jeff Saxe In SGI's online books I found: Arrays that are allocated on the process stack must not exceed 2 GB, but the total of all stack variables can exceed that limit. Apparently your routine is exceeding this limit. I get the same message on my SGI. However if I change from double precision to single precision it will compile. To get it to compile using double precision you will have to decrease your dimensions from 256 to about 160. I also have a suggestion concerning your FORTRAN program. Consider using parameter statements. I'd write the subroutine as program test IMPLICIT DOUBLE PRECISION (A-H,O-Z) parameter (nn = 256) DOUBLE COMPLEX AA(nn,nn,nn) DIMENSION ID(64),CC(64) DIMENSION NORD(3) DO 100 i=1,nn DO 100 j=1,nn DO 100 k=1,nn xx=DBLE(I+j+k) AA(i,j,k)=DCMPLX(xx,0.0D0) 100 CONTINUE NORD(1)=8 NORD(2)=8 NORD(3)=8 IERR=0 CALL DFFT3D(dta,NORD,ID,CC,1,IERR) STOP END This way, should you have to change the size, you simply change the parameter statement, and all the dimension statements and DO loops are automatically properly set. It avoids the possibility of forgetting to change something should you make a change. Another suggestion, though this is very picky and likely unnecessary these days, is to not end multiple DO loops with the same CONTINUE statement. I would have DO 100 DO 110 DO 120 ... 120 CONTINUE 110 CONTINUE 100 CONTINUE My only reason for this is that I have seen compilers make mistakes with this in the past, e.g. when the innermost loop in your code finished, it made the mistake of saying the outermost loop was done. Admittedly this was long ago (I've been writing FORTRAN for 22 years) and unlikely to happen now, but I still don't take chances myself. Jeff From: "Jochen Küpper" Y Z writes: > It is not because of FFT subroutine. It is > because of the large array size that I meet. > Anyway, I will try to use three one dimensional > FFT instead of a 3D-FFT. That will decrease the > speed of my code. Well, I don't see a problem in _that_ ? 256^3 =3D 2^{24} =3D 16777216 sizeof( double complex ) =3D 16 (I assume) =3D> 256 MB of RAM needed for the array If you need that amount of RAM, that's no problem at all (well, besides the money, probably :-). Anyway, it should work even in "swap space", although very slow and a better algorithm (memory wise) would do much better. But I don't=20= see any reason it doesn't compile - besides you are working on (Win)Doze ??? And of course your code has been correct. (Actually I never really looked through your code, and don't have your=20= first mail anymore.) But then I was thinking of memory-optimized algorithms that might be implemented in FFTW or available at netlib ? Just to let you know - I have written programs that allocate GB's (GigaBytes) of main-memory. And they actually compile and run :-) Since your code was FORTRAN - the program might actually crash at run=20= time due to missing stack-memory - you then need to increase the stack=20= size for the program or the system - need to be admin for that :-) But it shouldn't affect compilation. Jochen From: wsteinmetz It appears that you have run out of memory. Have you tried a (128,128,128) matrix that will be smaller by a multiplicative factor of 8. that should fit. From: laidig@pg.com As I mentioned, if you are running the c shell (or tcsh) as your UNIX shell you should be able to type: limit stacksize newstacksize where newstacksize is an integer variable of the new byte size limit This should work on all SGI's. Check with the command: limit If you are running another shell I don't know the command to use. Also, if the new limit is larger than the maximum size then your system manager will have to up the kernel parameter for the max. of this value (I don't remember the variable name) and have to reboot. Hope this helps, Bill From: Aeint de Boer I would add the statement: SAVE AA to your program and give it another try. ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From chemistry-request@server.ccl.net Tue Feb 22 19:32:16 2000 Received: from hotmail.com (f62.law3.hotmail.com [209.185.241.62]) by server.ccl.net (8.8.7/8.8.7) with SMTP id TAA09300 for ; Tue, 22 Feb 2000 19:32:15 -0500 Received: (qmail 94360 invoked by uid 0); 22 Feb 2000 23:23:19 -0000 Message-ID: <20000222232319.94359.qmail@hotmail.com> Received: from 128.101.162.4 by www.hotmail.com with HTTP; Tue, 22 Feb 2000 15:23:19 PST X-Originating-IP: [128.101.162.4] From: "Y Z" To: CHEMISTRY@ccl.net Subject: Analytic solution of power series Date: Tue, 22 Feb 2000 18:23:19 EST Mime-Version: 1.0 Content-Type: text/plain; format=flowed Are there analytic solution fomulus of the power series: infinity Rx-x SUM [---------------------] ;(m is not equal to 1) n=0 -> -> m ABS( nR - r ) -> ^ ^ ^ r = x i + y j + z k ; is a vector of (x,y,z) -> ^ ^ ^ R = Rx i + Ry j + Rz k is another vector (Rx,Ry,Rz) OR: In FORTRAN fotmat: SUM(n=0, to infinity): (Rx-x)/SQRT((Rx-x)**2+(Ry-y)**2+(Rz-z)**2)**(m/2) (m.ne.2) Thanks in advance. ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From chemistry-request@server.ccl.net Wed Feb 23 15:28:36 2000 Received: from ccl.net (atlantis.ccl.net [192.148.249.4]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id PAA16893 for ; Wed, 23 Feb 2000 15:28:35 -0500 Received: from dybbuk.hellers.com (dybbuk.hellers.com [207.197.221.3]) by ccl.net (8.8.6/8.8.6/OSC 1.1) with ESMTP id OAA15351 for ; Wed, 23 Feb 2000 14:21:19 -0500 (EST) Received: from localhost (steve@localhost) by dybbuk.hellers.com (8.8.8/8.8.7) with SMTP id OAA16184; Wed, 23 Feb 2000 14:33:27 -0500 (EST) Date: Wed, 23 Feb 2000 14:33:26 -0500 (EST) From: "Dr. Stephen R. Heller" X-Sender: steve@dybbuk Reply-To: "Dr. Stephen R. Heller" To: chemistry@ccl.net, chemweb@ic.ac.uk, chminf-l@listserv.indiana.edu, orgchem@extreme.chem.rpi.edu Subject: Software/Database for Review Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII February 22, 2000 Subject: Computer Software/Database for Review As the Software Review Editor for the ACS Journal of Chemical Information and Computer Science (JCICS) I often get software for review in the journal. I have one (1) new web-based software/database product. I am looking for someone who is willing to review this product on the web. The review should be completed in 1-3 months. The length of the review is 4-10 double spaced typed pages. Sample reviews can be found in most of the recent issues of JCICS. Please try to give me some (short) reason to choose you over another person. I have tried this approach for about the past five years and it is working reasonably well. (REMINDER: For those who haven't finished your reviews of software sent months and months ago, this last sentence does not apply to you!) As a result, I am continuing this new method to find reviewers using this e-mail/user group system. I reserve the right to abandon this if it is a problem, or inappropriate. I will not notify people if I have found a reviewer. If you don't hear from me within a few days I have chosen someone else to review the particular package. As I get many, many, (too many) replies to this message, please do not respond after 25 February 2000 (Friday), as I am sure the software will be gone by then. I can be reached on Internet (STEVE@HELLERS.COM). This is a new e-mail address. Steve Heller The package I now have is: ACD/I-Lab, the Interactive Laboratory is an online service including systematic naming generation, NMR and physicochemical prediction, and online database searching by structure and substructure. The person doing the review will be given a few hundred dollars credit to to access www.acdlabs.com/ilab to try it out and then write the review. While it is an online service but comes connected to the FREE Chemsketch interface for people to connect to. The reviewer would need to download CHemsketch to connect or use the Java applet. The reviewer MUST be using a very recent version of Internet Explorer or Netscape Navigator. Stephen R. Heller, Ph. D. Guest Researcher NIST/SRD, Mail Stop: 820/113 100 Bureau Drive 820 Diamond Avenue, Room 101 Gaithersburg, MD 20899-2310 USA Phone: 301-975-3338 FAX: 301-926-0416 E-mail: steve@hellers.com WWW: www.hellers.com/~steve -------------------------------------------------- As a member of the Organizing Committee, I invite you to attend Chemistry & the Internet - ChemInt2000; September 23-26, 2000; Washington, DC. http://www.chemint.org From chemistry-request@server.ccl.net Wed Feb 23 16:17:03 2000 Received: from yogi.pc1.uni-duesseldorf.de (yogi.pc1.uni-duesseldorf.de [134.99.152.44]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id QAA17396 for ; Wed, 23 Feb 2000 16:17:03 -0500 Received: (from jochen@localhost) by yogi.pc1.uni-duesseldorf.de (8.9.3/8.9.3) id VAA08652; Wed, 23 Feb 2000 21:09:49 +0100 From: "Jochen Küpper" MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Message-ID: <14516.16009.150855.561517@yogi.pc1.uni-duesseldorf.de> Date: Wed, 23 Feb 2000 21:09:44 +0100 (CET) To: Stefan Seidler Cc: jmmckel@attglobal.net, chemistry@ccl.net Subject: CCL:Error Function [ERF] Code (also [FAQ] :-) In-Reply-To: References: <38B256A8.9AA0BCCF@attglobal.net> X-Mailer: VM 6.72 under 21.1 (patch 4) "Arches" XEmacs Lucid Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by server.ccl.net id QAA17397 Stefan Seidler writes: > On Tue, 22 Feb 2000 jmmckel@attglobal.net wrote: > Look at Chapter 6.2 in Numerical Recipes, available online: > > http://www.ulib.org/webRoot/Books/Numerical_Recipes/bookfpdf/f6-2.pdf And look at http://math.jpl.nasa.gov/nr/nr-alt.html as well. FAQ: Q: What is about his Numerical Recipes routines ? A: Read the book, understand the routine, then go ahead and find a better implemetation at netlib or ... . NR is great for educational purposes - no more, no less. That's actually what the book is written for, isnßt it. Jochen -- Heinrich-Heine-Universität Institut für Physikalische Chemie I Jochen Küpper Universitätsstr. 1, Geb. 26.43.02.29 40225 Düsseldorf, Germany phone ++49-211-8113681, fax ++49-211-8115195 http://www.jochen-kuepper.de From chemistry-request@server.ccl.net Wed Feb 23 18:47:03 2000 Received: from ivory.trentu.ca (trentu.ca [192.75.12.103]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id SAA18790 for ; Wed, 23 Feb 2000 18:47:02 -0500 Received: from trentu.ca ([204.225.13.50]) by trentu.ca (PMDF V5.2-32 #29543) with ESMTP id <01JM97C6IZ6A000XWM@trentu.ca> for chemistry@ccl.net; Wed, 23 Feb 2000 17:39:41 EST Date: Wed, 23 Feb 2000 17:43:25 -0500 From: elewars Subject: SAM1 semiempirical method To: chemistry@ccl.net Message-id: <38B4628D.F1551EB0@trentu.ca> MIME-version: 1.0 X-Mailer: Mozilla 4.7 [en] (WinNT; I) Content-type: text/plain; charset=us-ascii Content-transfer-encoding: 7bit X-Accept-Language: en Wed, 2000 Feb 23 Hello, In 1993 the Dewar group announced the development of the SAM1 ("semi ab initio #1") method. I know the basic principles behind it, but would like to know: 1) What became of it? It does not seem to appear much in the literature. 2) Did computational chemists conclude it is not much better than AM1, and so let it fall into disuse? 3) I think it is offered in AMPAC; does any other package have it? Thanks EL ==== From chemistry-request@server.ccl.net Wed Feb 23 17:32:17 2000 Received: from relay1.scripps.edu (relay1.scripps.edu [137.131.200.29]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id RAA18213 for ; Wed, 23 Feb 2000 17:32:16 -0500 Received: from redox.scripps.edu (redox.scripps.edu [137.131.240.50]) by relay1.scripps.edu (8.9.3/TSRI-3.1.0r) with ESMTP id NAA20872; Wed, 23 Feb 2000 13:23:04 -0800 (PST) Received: from localhost (jesusmc@localhost) by redox.scripps.edu (8.9.2/TSRI-3.0.1) with SMTP id NAA05423; Wed, 23 Feb 2000 13:23:17 -0800 (PST) Date: Wed, 23 Feb 2000 13:23:17 -0800 (PST) From: "Jesus M. Castagnetto" Sender: jesusmc@scripps.edu To: Jan Labanowski cc: chemistry@ccl.net Subject: Re: CCL:FAQ and Search on CCL In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Jan's points on time and resources are correct, and his ideas on contributions to the FAQ the way to go. I know of two possible ways of automating the collection of FAQ questions and answers: 1) Using a FAQ-O-Matic system. There are systems like this that use Perl/Python/Java/PHP with and without a database backend, so people can put a FAQ question and/or answer, and comments can later be added by other people in the list. The *canonical* FAQ-O-Matic is at: http://www.dartmouth.edu/~jonh/ff-serve/cache/1.html and uses Perl and rcs to kep track of changes. The FAQ-O-Matic takes a hierarchical approach and has its own search scripts. There are ports of this package to other scripting languages. 2) A hosted knowledge base approach. From the point of view of the user it will be similar to (1), but the logistics in the server side are different. Also, this not necessarily needs to be hosted/maintained by Jan. There are places that can host a FAQ/knowledge base, and Jan can be the moderator, with the capacity to remove improper postings. For an example of (2), see the URL: http://www.faqts.com/knowledge-base/index.phtml/fid/51/ which is a knowledge base for PHP and related web stuff, and it is hoste at FAQTS.com, which also allow the creation of similar knowledge bases for free, see: http://www.faqts.com/index.phtml If we had a system like this, then the FAQ creation will be a community effort, and will not tax (too much) our good friend Jan. Regards. On Wed, 23 Feb 2000, Jan Labanowski wrote: > Date: Wed, 23 Feb 2000 14:13:18 -0500 (EST) > From: Jan Labanowski > To: chemistry@ccl.net > Cc: Jan Labanowski > Subject: CCL:FAQ and Search on CCL > > Dear, > > As to CCL FAQ and searches... > > FAQ... I would welcome it as a community project... > My problem is with time and the staff. The students who help me with CCL > are computer science students, not the Computational Chemistry oriented. > So I cannot expect them to help me with Computational Chemistry. > > But you can... > > When You see a FAQ question, and you respond to it in more or less > "closed" form (i.e., not with a one liner, but something which will > help a person who is not a guru), please copy it to me (jkl@ccl.net) > with a note "Add to FAQ" or upload it to incoming of our > anonymous ftp site: > http://www.ccl.net/chemistry/aboutccl/contributing/ > After I have some critical mass of these FAQ topics/answers, I will put it > on the web. I may even create a web form for FAQ submissions later. > Please help. I am short handed so do not expect me to act as an > editor... Just put the stuff in a closed form, so I can pass it to > a student who does not know what it is, and he/she can edit it into > the FAQ tree. I really think that this would improve the quality of the list > and discussions in general. It would be best, if you could make these > contributions in HTML. All the links sould be also displayed, e.g.: > > http://www.ccl.net/chemistry > > The format can be as: > > Author: your name and e-mail address > > Date: the current date > > Topic: Keywords relevant to the topic below > > Question: The precise question to which you provide answer > > Answer: The gory details, but less than a few pages. If answer needs more > then few pages of narrative, it is not an answer to a Question, > it is a review (which I always welcome and you have some o > them under "Document Archives"). You can of course have > some examples or accompanying files for the FAQ, please attach > them. The other option is, make it a tar or zip, and upload it > to the incoming directory on ftp.ccl.net and let me know. > > If you have a better idea, send it to me, and I will choose the thing > which best fits the overall structure of CCL. > Do not expect answers soon, since I will be at Sanibel for a week. > > SEARCHES.... > > Sorry... Our htdig search started to produce indices which want to be > larger than 2Gigs of space. Of course, a no, no under older Linices > which does not like files larger than 2Gigs (yes, we are upgrading slowly). > So, we have to find the solution. In the meantime use the Regular Expression > search. Sorry, it is slow (and often dies), but this is what we have at this > moment. My students are looking at alternatives/solutions. > > Thank you for your patience... > > Jan > jkl@ccl.net > > Jan K. Labanowski | phone: 614-292-9279, FAX: 614-292-7168 > Ohio Supercomputer Center | Internet: jkl@ccl.net > 1224 Kinnear Rd, | http://www.ccl.net/chemistry.html > Columbus, OH 43212-1163 | http://www.ccl.net/ > > > > -= This is automatically added to each message by mailing script =- > CHEMISTRY@ccl.net -- To Everybody | CHEMISTRY-REQUEST@ccl.net -- To Admins > MAILSERV@ccl.net -- HELP CHEMISTRY or HELP SEARCH > CHEMISTRY-SEARCH@ccl.net -- archive search | Gopher: gopher.ccl.net 70 > Ftp: ftp.ccl.net | WWW: http://www.ccl.net/chemistry/ | Jan: jkl@ccl.net > > > > > > -- Jesus M. Castagnetto - "Ken Zen Ichi-nyo" Program Project: http://www.scripps.edu/research/metallo/ Metalloprotein DB: http://metallo.scripps.edu/ Pilot Stuff: http://www.geocities.com/ResearchTriangle/Lab/1059/ From chemistry-request@server.ccl.net Wed Feb 23 22:23:24 2000 Received: from schrodinger.com (root@thermidore.schrodinger.com [192.156.98.99]) by server.ccl.net (8.8.7/8.8.7) with ESMTP id WAA19748 for ; Wed, 23 Feb 2000 22:23:23 -0500 Received: from sally.schrodinger.com (shenkin@sally.schrodinger.com [166.84.149.25]) by schrodinger.com (8.8.5/8.8.5) with ESMTP id SAA26696 for ; Wed, 23 Feb 2000 18:16:13 -0800 Received: from localhost (shenkin@localhost) by sally.schrodinger.com (8.8.8/8.8.5) with ESMTP id VAA09618 for ; Wed, 23 Feb 2000 21:16:06 -0500 X-Authentication-Warning: sally.schrodinger.com: shenkin owned process doing -bs Date: Wed, 23 Feb 2000 21:16:06 -0500 (EST) From: Peter Shenkin To: chemistry@ccl.net Subject: Re: CCL:The "birth-year" of computational chemistry? In-Reply-To: <200002231458.JAA51131@phu847.um.us.sbphrd.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi, Dominic, When I was in high-school, I built a suspended ball-spring version of the Kettering device as a sort of senior project. This brings back memories. But I don't think you'd call this computational chemistry. :-) I think that term should be reserved for the use of digital computers. In this regard, I believe the Metropolis paper was in 1948, and might have been the first. -P. On Wed, 23 Feb 2000, M Dominic Ryan wrote: > I point back to early spectroscopic work. The earliest I think is work by > Andrews, in the 20s. Once citation I have is in Physical Reviews, v36, > p531, 1930, Kettering, Shutts and Andrews (General Motors Corporation). > "A representation of the Dynamic Properties of Molecules by Mechanical > Models". I think of this as the birth of molecular modeling. They built > molecular models of things like toluene, benzene, chlorobenzene using > steel balls and calibrated springs. -- ** Whether the playing field is level depends on the coordinate system. *** ********* Peter S. Shenkin; Schrodinger, Inc.; (201)433-2014 x111 ********* *********** shenkin@schrodinger.com; http://www.schrodinger.com ***********