simulation@uflorida.cis.ufl.edu (Moderator: Paul Fishwick) (10/24/89)
Volume: 11, Issue: 10, Tue Oct 24 09:53:22 EDT 1989 +----------------+ | TODAY'S TOPICS | +----------------+ (1) Random Number Generators (2) Availability of ROSS? (3) History of Simulation (4) Continuous Simulation Software * Moderator: Paul Fishwick, Univ. of Florida * Send topical mail to: simulation@bikini.cis.ufl.edu OR post to comp.simulation via USENET * Archives available via FTP to bikini.cis.ufl.edu, login as 'ftp', use your last name as the password, change directory to pub/simdigest. * Simulation Tools available by doing above and changing the directory to pub/simdigest/tools. ----------------------------------------------------------------------------- To: comp-simulation@ucsd.edu Date: Thu Oct 19 16:57:23 1989 From: garym@crash.cts.com (Gary Morris) To: ucsd!comp-simulation Path: crash!garym From: garym@crash.cts.com (Gary Morris) Newsgroups: comp.simulation,sci.math.stat,rec.games.programmer Subject: request for Random Number Generator algorithms and theory Summary: need references Keywords: random, algorithm, request Date: 19 Oct 89 23:57:22 GMT Reply-To: garym@crash.cts.com (Gary Morris) Followup-To: sci.math.stat Distribution: na Organization: Crash TimeSharing, El Cajon, CA Howdy! I need to write two different random number generators (RNG) into a simulation that I am writing. Presently I am doing research on the various algorithms for generating random numbers and for testing RNGs. I'm writing the simulation on a Macintosh so I can't use the random() or rand() function in the Unix library. The Random() function on the Mac only provides a 16 bit result and no information is given on how this is produced, so I can't judge if it is what I need, though it seems to give good chi-square figures. First I need the generator to give numbers that are uniformly distributed over the full range (which should be a minimum of 16 bits, ie: -32768..32767, but preferably 32 bits). It should also have a very long period as it will be used tens of millions of times during the simulation. The random function (random (3)) on our Unix system (Sun 4.0) sounds good according to the manual page but there is not much info on the implementation nor are references provided. The man page says: > DESCRIPTION > random uses a non-linear additive feedback random number > generator employing a default table of size 31 long integers > to return successive pseudo-random numbers in the range from > 0 to (2**31)-1. The period of this random number generator > is very large, approximately 16*((2**31)-1). Second, I want to be able to simulate a selection. I'm not sure if I need a special RNG for this or if I can filter a uniformly distributed RNG to get a weighted distribution. Consider a simulation of someone inserting the cut card into the middle of a deck of cards. Assume the person is attempting to insert the card into the middle of the deck. If I were to use a random number to pick the insertion point, it would be evenly distributed over the range 1..52. In reality, the distribution should be heavily weighted to the center of the deck with the frequency of occurrence dropping away from the center, since someone trying to insert in the center will usually get pretty close and rarely insert near the edge. Hear's a graph that attempts to show what I'm trying to do (vertical axis is frequency of occurrence, horizontal is the position within the deck or random value): | | | _ | Freq: | | | / \ | | | | / \ | |-----------------| | / \ | | | | / \ | | | | / \ | +-----------------+ +---------------+ Value: 1 26 52 1 26 52 uniform distribution weighted to center If you have information, algorithms or references to publications that could provide solutions to these questions I would very much appreciate if you would email a note to me (addresses below). Please don't respond directly to this article (I don't always get all the groups this is posted to), please send email to one of the addresses in my signature below. If there is interest, I could post a summary of my research. Thanks!! --GaryM Gary Morris {uunet,ucsd}!telesoft!garym TeleSoft N6FRT garym@telesoft.com San Diego, CA, USA ASEL garym@crash.cts.com +1 619 457-2700 "If something is worth doing, it's worth doing right." -- Gary Morris uunet!ucsd!telesoft!garym TeleSoft, San Diego telesoft!garym@ucsd.ucsd.edu (619) 457-2700 garym@crash.cts.com ------------------------------ Date: 20 Oct 89 11:24:21 GMT From: Schneider Daniel <shneider@cui.unige.ch> To: <fishwick@bikini.cis.ufl.edu> Subject: The simulation language ROSS ? I am looking for a pointer to the simulation language ROSS which runs under SCheme and seems to useful for qualitative simulations. Would anyone know by any chance where I can get more information about ROSS ? Or do you know of any other other language written Lisp or Scheme for which source code is available? I know about Sim-kit, but it is too big and much to expensive for my current purpose. - thanks for helping me out if you can - Daniel Schneider Daniel K.Schneider, Faculte de Psychologie et des Sciences de l'Education, University of Geneva, 1211 Geneva 4 (Switzerland), Tel. (..41) (22) 705 7652 Internet: shneider@cui.unige.ch (and various national nets) | if reply does CSnet/ARPA: shneider%cui.unige.ch@relay.cs.net (old style) | not work, JANET: shneider%cui.unige.ch@uk.ac.ean-relay | try one of uucp: mcvax!cernvax!cui!shneider | these BITNET: SCHNEIDE@CGEUGE51 DECNET: UGUN2A::SCHNEIDE (local) ------------------------------ Posted-Date: Fri, 20 Oct 89 13:49:43 CDT Date: Fri, 20 Oct 89 13:49:43 CDT From: steve@titan.tsd.arlut.utexas.edu (Steve Glicker) To: simulation@bikini.cis.ufl.edu Subject: History of simulation A few weeks ago I posted a request to the Simulation Digest for references to accounts of the History of Simulation. I have received a few interesting leads: o "Pioneers and Peers" edited by John McLeod and sold by the SCS in San Diego (from Paul Fishwick), o Historical articles on GPSS and Simula in the proceedings of ACM's History of Programming Languages Conference held in 1978 (from Ron Kerr and Mike Lemon) and the reference in the article recently posted by A. S. Elmaghraby to this newsgroup. (Thanks for the help!) I've recently come across a section in Wolfgang Kreutzer's book (1986) "System Simulation Programming Styles and Languages" entitled "History and family lines" where he provides two family trees for continuous and combined simulation languages as well as for discrete languages. Furthermore, he writes, The IFIP conference on Simulation Programming Languages in 1968 [Buxton (1968)] was a major milestone in the history of discrete-event simulation. Since then the development of new systems has slowed, but there has been some convergence of the most important language families. ... His reference, which I have not yet seen, is: Buxton, J.N. (ed.) (1968). "Simulation Programming Languages". Proceedings of the IFIP Working Conference on Simulation Programming Languages, Amsterdam/New York: North-Holland. I consider myself to be in the early stage of gathering information on this topic and would greatly appreciate any leads. Steve Glicker (steve@titan.tsd.arlut.utexas.edu) ------------------------------ Date: Tue, 24 Oct 89 01:30:59 EDT From: Rajeev Jain <jainr@clutx.clarkson.edu> To: simulation@bikini.cis.ufl.edu Subject: simulation software Cc: jainr@clutx.clarkson.edu I am devoloping an ecological model for Lake Ontario and would like to know about various simulation software available for continously time variable systems. My system is large (about 600 state variables) , highly interlinked, reasonably nonlinear and likely to be stiff. As of today, I am aware of only one package: ACSL. Would anybody make any recommendations for the above kind of system? I would appreciate any general comments or specific comments on chemical-ecological models. Rajeev. ------------------------------ END OF SIMULATION DIGEST ************************