From wajnberg at sophia.inra.fr Thu Apr 23 04:10:34 2009 From: wajnberg at sophia.inra.fr (Eric Wajnberg) Date: Thu Apr 23 04:10:36 2009 Subject: seed for random number generator with parallel computation Message-ID: Dear all, For years I am developing my codes to run Monte Carlo simulations. For this I am using the random number generator ran2 from L'Ecuyer (numerical recipes) without any problem. Actually, in order to have a continuous flow of random values, each run of my codes stores at the end a seed value in a file that is then read on next run to initiate the random number generator, etc. The point is that the running time for the problem I have to solve now is far too long and I thus decided to implement my codes on a cluster. The computation time obviously drastically reduced but I am now facing a new problem that I cannot solve. The point is that several runs of my codes are now launched simultaneously, and are thus using the same seed, which is then wrong. Can someone tell me what is the usual policy used in this case to start the random number generator with random seeds, that are different from one run to the other? Are you simply using the pid of the process each time to initiate the random number generator? Any hint on this will be welcomed. Thanks in advance for that. Cheers, Eric. From chris at csse.uwa.edu.au Thu Apr 23 05:38:34 2009 From: chris at csse.uwa.edu.au (Chris McDonald) Date: Thu Apr 23 05:38:36 2009 Subject: seed for random number generator with parallel computation References: Message-ID: Eric Wajnberg writes: >Can someone tell me what is the usual policy used in this case to start >the random number generator with random seeds, that are different from >one run to the other? Are you simply using the pid of the process each >time to initiate the random number generator? There's not many different pids available, some Unix systems will give you only 30,000 of them. If your system provides the pseudo-device /dev/random, you may like to investigate it: man 4 random -- Chris, From jdsalt_AT_gotadsl.co.uk at giganews.com Thu Apr 23 13:25:18 2009 From: jdsalt_AT_gotadsl.co.uk at giganews.com (John D Salt) Date: Thu Apr 23 13:25:21 2009 Subject: seed for random number generator with parallel computation References: Message-ID: Eric Wajnberg wrote in news:gsp7ns$ems$1 @anthracite.jouy.inra.fr: [Snips] > Can someone tell me what is the usual policy used in this case to start > the random number generator with random seeds, that are different from > one run to the other? Are you simply using the pid of the process each > time to initiate the random number generator? I don't know that it's "usual policy", but my habit is to write a little standalone program to make a large number of calls to the PRNG, and record the current state of the generator at regular intervals. Obviously the exact numbers depend on the number of random draws you expect to make in a replication, and the period of the generator, but I tend to make a million random draws and record the seed at intervals of 10,000, so giving me in effect 100 starting seeds to choose from. In the event that you ever draw more than 10,000 samples from a single random stream, I suspect that any correlation effects with the random stream it is now overlapping will be negligible. I originally did this with CACI's SIMSCRIPT II.5 and MODSIM II, in order to provide more than the 10 starting seeds provided "out of the box"; I have since done it with Python, even though the Python PRNG provides a method called jumpahead() to move far away from the current position in the cycle. My reason is that I prefer to know the exact definition of "far away" for these purposes. I certainly wouldn't use a process pid as a PRNG seed, for fear of losing repeatability. As an aside, starting with known seeds in the way suggested offers you the chance to use antithetic variates as a variance- reduction technique, if you care to do so. All the best, John. -- In what method shall we implement the matrix of this government display picnic? -- Bill Bailey From YuriyMikhaylovskiy at gmail.com Thu Apr 23 18:46:52 2009 From: YuriyMikhaylovskiy at gmail.com (Yuriy) Date: Thu Apr 23 18:46:54 2009 Subject: Need help in the development and use JMCAD References: <4025cf5e-0f89-4e56-8865-ac157e1da98a@h5g2000yqh.googlegroups.com> Message-ID: <1b70fb58-c994-4cd8-a038-9d19813a193e@c36g2000yqn.googlegroups.com> New version JMCAD-07.083 released!!! From newsgroup at wolfram.com Wed Apr 29 18:51:29 2009 From: newsgroup at wolfram.com (newsgroup@wolfram.com) Date: Wed Apr 29 18:51:31 2009 Subject: Introducing the Wolfram Mathematica Tutorial Collection Message-ID: <259c0372-efb8-4e0c-a991-3f517b0c93c7@v35g2000pro.googlegroups.com> Our users asked for it, and we've created it. The much anticipated Wolfram Mathematica Tutorial Collection is now available in a variety of formats to suit your needs: on the web, as PDFs, and, by popular demand, in print. The Tutorial Collection gives you targeted instruction with full information on the functions, capabilities, and unified architecture of the Mathematica system, all built into one anthology. Drawn from the in-product documentation of Mathematica, the collection includes 23 volumes totaling nearly 3700 pages. Topics include core language, mathematics and algorithms, visualization and graphics, notebooks and documents, dynamic interactivity, data manipulation, and systems interfaces and deployment. For more information, visit: http://www.wolfram.com/tutorialcollection Sincerely, Maryka Baraka Author and Publisher Support Specialist Wolfram Research, Inc.