C++Talk.NET C++ language newsgroups   Archives   FAQ   Search   Memberlist   Usergroups   Register   Profile   Log in to check your private messages   Log in  Randomizer by time. Goto page 1, 2  Next          C++Talk.NET Forum Index -> C++ Language (Moderated) View previous topic :: View next topic   Author Message kurdayon@yahoo.com Guest Posted: Wed Jan 19, 2005 11:18 pm    Post subject: Randomizer by time. Hello, to generate random integer number from a given range I use the following lines: srand(time(0)); tmp = Nmin + rand() % Nmax; However, if interval between subsequent call of these lines is small I get the same numbers. The cause is clear (argument time in strand is the same). Not clear is how I can resolve this problem. Can you help me? [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Maciej Sobczak Guest Posted: Thu Jan 20, 2005 11:51 am    Post subject: Re: Randomizer by time. Hi, [email]kurdayon (AT) yahoo (DOT) com[/email] wrote: Quote: Hello, to generate random integer number from a given range I use the following lines: srand(time(0)); tmp = Nmin + rand() % Nmax; However, if interval between subsequent call of these lines is small I get the same numbers. The cause is clear (argument time in strand is the same). Not clear is how I can resolve this problem. Can you help me? Just do not clump these two lines together. Normally, I would expect srand() to be called *once* (for example, at the beginning of the program or anytime before the first rand() is called). After doing that, call only rand(), as many times as you want. Another issue is that your formula for "random" number from a given range may not give a uniform distribution, depending on the relation between Nmax and RAND_MAX. See also: http://www.eskimo.com/~scs/C-faq/q13.16.html -- Maciej Sobczak : http://www.msobczak.com/ Programming : http://www.msobczak.com/prog/ [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Jack Klein Guest Posted: Thu Jan 20, 2005 11:52 am    Post subject: Re: Randomizer by time. On 19 Jan 2005 18:18:47 -0500, [email]kurdayon (AT) yahoo (DOT) com[/email] wrote in comp.lang.c++.moderated: Quote: Hello, to generate random integer number from a given range I use the following lines: srand(time(0)); Note that this code does not necessarily have defined behavior, depending on the type used for size_t by the implementation and the value that size_t might contain. Quote: tmp = Nmin + rand() % Nmax; However, if interval between subsequent call of these lines is small I get the same numbers. The cause is clear (argument time in strand is the same). Not clear is how I can resolve this problem. Can you help me? It is a very common newbie mistake to call srand() before every call to rand(). There is almost never an actual reason to call srand() more than once. In general, call srand() once only at the start of the program, then call rand() as often as you like. See also the questions about random numbers in section 13 of the FAQ for comp.lang.c, here http://www.eskimo.com/~scs/C-faq/s13.html. They apply equally well in C++. -- Jack Klein Home: http://JK-Technology.Com FAQs for comp.lang.c http://www.eskimo.com/~scs/C-faq/top.html comp.lang.c++ http://www.parashift.com/c++-faq-lite/ alt.comp.lang.learn.c-c++ http://www.contrib.andrew.cmu.edu/~ajo/docs/FAQ-acllc.html [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Ian McCulloch Guest Posted: Thu Jan 20, 2005 12:18 pm    Post subject: Re: Randomizer by time. [email]kurdayon (AT) yahoo (DOT) com[/email] wrote: Quote: Hello, to generate random integer number from a given range I use the following lines: srand(time(0)); tmp = Nmin + rand() % Nmax; However, if interval between subsequent call of these lines is small I get the same numbers. The cause is clear (argument time in strand is the same). Not clear is how I can resolve this problem. Can you help me? You should initialize the random seed just once at the start of the program and then leave it. rand() will return a more-or-less random sequence without further intervention. Setting the seed more frequently is pointless, and in this case - where the seed has practically zero entropy - will destroy whatever randomness properies your implementation of rand() has anyway. HTH, Ian [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top jakacki Guest Posted: Thu Jan 20, 2005 12:19 pm    Post subject: Re: Randomizer by time. Quote: to generate random integer number from a given range I use the following lines: srand(time(0)); tmp = Nmin + rand() % Nmax; However, if interval between subsequent call of these lines is small I get the same numbers. The cause is clear (argument time in strand is the same). Not clear is how I can resolve this problem. Can you help me? Call 'srand()' only once, then obtain consecutive numbers by calling just 'rand()'. BR Grzegorz -- Free C++ frontend library: http://opencxx.sourceforge.net China from the inside: http://www.staryhutong.com Myself: http://www.dziupla.net/gj/cv [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Efrat Regev Guest Posted: Thu Jan 20, 2005 12:23 pm    Post subject: Re: Randomizer by time. Hello, The boost libraries contain very good random-number classes by Jens Maurer. One of the things you can see is that you can create a randomization engine object, and use it for random number generators. The possible relevance for you is that, in this design, you only need to "srand" only a "global" engine object. Consequently, ::time(0) will be called only once per program execution, and, therefore, subsequent calls to ::time(0) will be far apart in time. I hope this helps <kurdayon (AT) yahoo (DOT) com> wrote Quote: Hello, to generate random integer number from a given range I use the following lines: srand(time(0)); tmp = Nmin + rand() % Nmax; However, if interval between subsequent call of these lines is small I get the same numbers. The cause is clear (argument time in strand is the same). Not clear is how I can resolve this problem. Can you help me? [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top kanze@gabi-soft.fr Guest Posted: Thu Jan 20, 2005 6:17 pm    Post subject: Re: Randomizer by time. [email]kurdayon (AT) yahoo (DOT) com[/email] wrote: Quote: to generate random integer number from a given range I use the following lines: srand(time(0)); tmp = Nmin + rand() % Nmax; However, if interval between subsequent call of these lines is small I get the same numbers. The cause is clear (argument time in strand is the same). Not clear is how I can resolve this problem. Can you help me? You're only supposed to call srand() once, at the beginning of the program. -- James Kanze GABI Software http://www.gabi-soft.fr Conseils en informatique orientée objet/ Beratung in objektorientierter Datenverarbeitung 9 place Sémard, 78210 St.-Cyr-l'École, France, +33 (0)1 30 23 00 34 [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Peter Koch Larsen Guest Posted: Thu Jan 20, 2005 6:22 pm    Post subject: Re: Randomizer by time. <kurdayon (AT) yahoo (DOT) com> skrev i en meddelelse news:1106164595.205998.84760 (AT) c13g2000cwb (DOT) googlegroups.com... Quote: Hello, to generate random integer number from a given range I use the following lines: srand(time(0)); tmp = Nmin + rand() % Nmax; However, if interval between subsequent call of these lines is small I get the same numbers. The cause is clear (argument time in strand is the same). Not clear is how I can resolve this problem. Can you help me? [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] You should call srand only once in your program. /Peter [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Joseph Turian Guest Posted: Fri Jan 21, 2005 10:27 am    Post subject: Re: Randomizer by time. Why are you using rand() instead of the rand48 functions? You'll get much better random numbers that way. Joseph [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Seungbeom Kim Guest Posted: Sat Jan 22, 2005 4:54 am    Post subject: Re: Randomizer by time. Joseph Turian wrote: Quote: Why are you using rand() instead of the rand48 functions? You'll get much better random numbers that way. Probably because they are not standard and their availability depends on the environment..? -- Seungbeom Kim [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Francis Glassborow Guest Posted: Sat Jan 22, 2005 5:00 am    Post subject: Re: Randomizer by time. In article <1106263173.522428.185130 (AT) z14g2000cwz (DOT) googlegroups.com>, Joseph Turian <turian (AT) gmail (DOT) com> writes Quote: Why are you using rand() instead of the rand48 functions? You'll get much better random numbers that way. But not much use if you are trying to write portable C++, and not much use if you are using an implementation where long is only 32 bits. -- Francis Glassborow ACCU Author of 'You Can Do It!' see http://www.spellen.org/youcandoit For project ideas and contributions: http://www.spellen.org/youcandoit/projects [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top kanze@gabi-soft.fr Guest Posted: Sat Jan 22, 2005 5:33 am    Post subject: Re: Randomizer by time. Jack Klein wrote: Quote: On 19 Jan 2005 18:18:47 -0500, [email]kurdayon (AT) yahoo (DOT) com[/email] wrote in comp.lang.c++.moderated: to generate random integer number from a given range I use the following lines: srand(time(0)); Note that this code does not necessarily have defined behavior, depending on the type used for size_t by the implementation and the value that size_t might contain. Technically, no, although it may not do anything useful. (I am supposing that your size_t is a typo for time_t. If not, I don't understand; there's not a single size_t in sight.) If the necessary headers have not been included, then of course, it has undefined behavior. Always. If they have been included, then there should be no problem -- the time_t returned by time() is an arithmetic type, which can be implicitely converted to the unsigned int which srand expects. If the value cannot be represented as an unsigned int, then the results of the conversion are implementation defined (and may result in an implementation defined signal, at least in C99); implementation defined, and not undefined. Of course, on a 16 bit machine, with a Unix-like time_t, you probably will get an implementation defined signal, if that is what the implementation does. And if the implementation represents time in megayears since 1900, on a double, you're not likely to get a useful seed, even without the signal:-). But it is a consacrated idiom:-). -- James Kanze GABI Software http://www.gabi-soft.fr Conseils en informatique orientée objet/ Beratung in objektorientierter Datenverarbeitung 9 place Sémard, 78210 St.-Cyr-l'École, France, +33 (0)1 30 23 00 34 [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top kanze@gabi-soft.fr Guest Posted: Sat Jan 22, 2005 5:36 am    Post subject: Re: Randomizer by time. Joseph Turian wrote: Quote: Why are you using rand() instead of the rand48 functions? You'll get much better random numbers that way. First, of course, because rand48 is not portable. It's not part of Unix and it isn't present on Windows. There is a drand48 in the Open System extensions, but it's not Posix, and it isn't present on Windows either. Generally speaking, I think that the ?rand48 functions were introduced in Berkley Unix, because the rand() in the original Unix library was so bad. The actual function specified in the Open System's standards doesn't look all that good to me, and a lot of the rand() today have improved, so I'm not sure that ?rand48 is necessarily that much better. If you want to be sure of the quality of your random number generator, the only solution is to adopt a single implementation of known quality. Boost::random offers a lot, for example -- if you can't find what you need there, I don't know where you'll find it. Note too that just taking the modulo of the results will introduce some skewing toward the lower values. -- James Kanze GABI Software http://www.gabi-soft.fr Conseils en informatique orientée objet/ Beratung in objektorientierter Datenverarbeitung 9 place Sémard, 78210 St.-Cyr-l'École, France, +33 (0)1 30 23 00 34 [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Joseph Turian Guest Posted: Sun Jan 23, 2005 2:40 am    Post subject: Re: Randomizer by time. Seungbeom Kim writes: Quote: Probably because they are not standard and their availability depends on the environment..? Fair enough. Francis Glassborow writes: Quote: not much use if you are using an implementation where long is only 32 bits. Uh... why not? Quote: From the drand48 manpage on Linux: The lrand48() and nrand48() functions return non-negative long integers uniformly distributed between 0 and 2^31. The mrand48() and jrand48() functions return signed long integers uni- formly distributed between -2^31 and 2^31. Alternately: unsigned long myrand(unsigned long ceiling) { return (unsigned long)(drand48() * ceiling); } James Kanze writes: Quote: Note too that just taking the modulo of the results will introduce some skewing toward the lower values. True, depending upon your implementation. And since Seungbeom seems to be interested in using rand() on platforms where the implementation might be poor, he should heed your warning. Quote: From the rand() manpage: In Numerical Recipes in C: The Art of Scientific Computing (William H. Press, Brian P. Flannery, Saul A. Teukolsky, William T. Vetterling; New York: Cambridge University Press, 1992 (2nd ed., p. 277)), the follow- ing comments are made: "If you want to generate a random integer between 1 and 10, you should always do it by using high-order bits, as in j=1+(int) (10.0*rand()/(RAND_MAX+1.0)); and never by anything resembling j=1+(rand() % 10); (which uses lower-order bits)." Joseph [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top kanze@gabi-soft.fr Guest Posted: Mon Jan 24, 2005 9:02 pm    Post subject: Re: Randomizer by time. Joseph Turian wrote: Quote: Seungbeom Kim writes: Probably because they are not standard and their availability depends on the environment..? Fair enough. Francis Glassborow writes: not much use if you are using an implementation where long is only 32 bits. Uh... why not? From the drand48 manpage on Linux: The lrand48() and nrand48() functions return non-negative long integers uniformly distributed between 0 and 2^31. The mrand48() and jrand48() functions return signed long integers uni- formly distributed between -2^31 and 2^31. Alternately: unsigned long myrand(unsigned long ceiling) { return (unsigned long)(drand48() * ceiling); } James Kanze writes: Note too that just taking the modulo of the results will introduce some skewing toward the lower values. True, depending upon your implementation. Depending on the value you are using for the modulo and the implementation. If the value is a divisor of RAND_MAX+1, there will be no skew. But since RAND_MAX+1 will be a prime number if you are using a good linear congruent generator (which are still by far the most common), there is a good chance that it isn't a divisor. And of course, if you use % 6 one time, and %8 the next, at least one of the two will manifest skew. Quote: And since Seungbeom seems to be interested in using rand() on platforms where the implementation might be poor, he should heed your warning. From the rand() manpage: In Numerical Recipes in C: The Art of Scientific Computing (William H. Press, Brian P. Flannery, Saul A. Teukolsky, William T. Vetterling; New York: Cambridge University Press, 1992 (2nd ed., p. 277)), the follow- ing comments are made: "If you want to generate a random integer between 1 and 10, you should always do it by using high-order bits, as in j=1+(int) (10.0*rand()/(RAND_MAX+1.0)); and never by anything resembling j=1+(rand() % 10); (which uses lower-order bits)." IMHO, that's not really good advice. If you have a bad generator, using the high order bits won't make it a good one; it just means that you won't see the weakness as easily. If you have a good generator, using the low order bits is fine. None of which, of course, changes the fact that the modulo introduces skew toward the lower values. Regardless of the generator used. -- James Kanze GABI Software http://www.gabi-soft.fr Conseils en informatique orientée objet/ Beratung in objektorientierter Datenverarbeitung 9 place Sémard, 78210 St.-Cyr-l'École, France, +33 (0)1 30 23 00 34 [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Display posts from previous: All Posts1 Day7 Days2 Weeks1 Month3 Months6 Months1 Year Oldest FirstNewest First         C++Talk.NET Forum Index -> C++ Language (Moderated) All times are GMT Goto page 1, 2  Next Page 1 of 2   Jump to: Select a forum C/C++----------------C LanguageC Language (Moderated)C++ Language (Moderated)C++ language (comp.lang.c++)C++ language, library and standards Non-English C/C++ Newsgroups----------------C++ (German)C++ (French)  You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum Powered by phpBB © 2001, 2006 phpBB Group SEO toolkit © 2004-2006 webmedic.