Programming Forums > Application Development > C

mutyalarao · Mar 2nd, 2007, 11:02 AM

Hi all,

How are random numbers generated , even using some functions supplied by the programming language? whats the logic behind the random number generation algorithm?

jim mcnamara · Mar 2nd, 2007, 11:12 AM

This is the Knuth PRNG (Pseudo-Random Number generator) that was used in glibc3.1 for Linux, it is reentrant:

(Toggle Plain Text)

int
rand_r (unsigned int *seed)
{
  unsigned int next = *seed;
  int result;

  next *= 1103515245;
  next += 12345;
  result = (unsigned int) (next / 65536) % 2048;

  next *= 1103515245;
  next += 12345;
  result <<= 10;
  result ^= (unsigned int) (next / 65536) % 1024;

  next *= 1103515245;
  next += 12345;
  result <<= 10;
  result ^= (unsigned int) (next / 65536) % 1024;

  *seed = next;

  return result;
}

Wizard1988 · Mar 2nd, 2007, 11:20 AM

One of our members has a good explanation

.
http://www.eternallyconfuzzled.com/a..._art_rand.aspx

Fall Back Son · Mar 26th, 2007, 1:47 AM

short version:

Random numbers aren't really random. They are either "seeded" by a number or by the system time. If seeded by a number, they use calculations to come up with the "random" numbers. If seeded by the system time, they use that to come up with the random numbers. The system time method is better in some cases because it is closer to true randomness.

DaWei · Mar 26th, 2007, 1:51 AM

Quote:

Random numbers aren't really random.

Yes, they are. PSEUDO-random numbers aren't really random. You might want to read the material provided in the preceding posts.

Fall Back Son · Mar 26th, 2007, 2:04 AM

I was clearly referring to 'random numbers' in C programming. Does that clear it up, or am I still incorrect?

Game_Ender · Mar 26th, 2007, 2:08 AM

DaWei's point is that it is possible with the right hardware and libraries to get true random number in C. So just because we are talking about C and random numbers does not mean they are always pseudo-random.

DaWei · Mar 26th, 2007, 2:10 AM

You are apparently referring to the PRNG commonly found in the run-time library. This does not preclude one from using C with true random numbers. I'm of the opinion that disseminating flakey information, whether by loose semantics or otherwise, is not a good idea. Again, I'd suggest you follow the earlier link for additional considerations involved in seeding the PRNG with the system's time.

Fall Back Son · Mar 26th, 2007, 2:16 AM

Can you show me such a code? Not saying it doesn't exist, but it seems that any code would not be truly random.

"I'm of the opinion that disseminating flakey information, whether by loose semantics or otherwise, is not a good idea."

edit: DaWei, I'm all for correct grammar as well, but couldn't you just tell me to shut the hell up if I don't know what I'm talking about?

Also:

I already read everything in the original link. I'm still of the opinion that the numbers you say are "truly random" are only more significantly random than others. While I understand the article's point about distribution, a perfect distribution does not necessarily indicate true randomness. Is that incorrect?

DaWei · Mar 26th, 2007, 9:40 AM

I didn't mention grammar.

The statement was made that C can use processes that generate truly random numbers. No one mentioned that this will happen with any PRNG, regardless of it's distribution.

My view on misinformation remains the same; my expression seemed preferable, at the time of that post, to telling you to just shut the fuck up, read the posts, and maybe resort to Google.

Mar 2nd, 2007, 11:02 AM	#1
mutyalarao Newbie Join Date: Aug 2005 Location: Bangalore,India Posts: 1 Rep Power: 0	Can anyone explain Random number generation Hi all, How are random numbers generated , even using some functions supplied by the programming language? whats the logic behind the random number generation algorithm? __________________ Regards, Mutyala Rao Aripaka, Bangalore, INDIA mutyalarao@gmail.com

Mar 2nd, 2007, 11:12 AM	#2
jim mcnamara Hobbyist Programmer Join Date: Jun 2005 Location: New Mexico Posts: 228 Rep Power: 4	This is the Knuth PRNG (Pseudo-Random Number generator) that was used in glibc3.1 for Linux, it is reentrant: (Toggle Plain Text) int rand_r (unsigned int seed) { unsigned int next = seed; int result; next = 1103515245; next += 12345; result = (unsigned int) (next / 65536) % 2048; next = 1103515245; next += 12345; result <<= 10; result ^= (unsigned int) (next / 65536) % 1024; next = 1103515245; next += 12345; result <<= 10; result ^= (unsigned int) (next / 65536) % 1024; seed = next; return result; } int rand_r (unsigned int seed) { unsigned int next = seed; int result; next = 1103515245; next += 12345; result = (unsigned int) (next / 65536) % 2048; next = 1103515245; next += 12345; result <<= 10; result ^= (unsigned int) (next / 65536) % 1024; next = 1103515245; next += 12345; result <<= 10; result ^= (unsigned int) (next / 65536) % 1024; seed = next; return result; }

Mar 2nd, 2007, 11:20 AM	#3
Wizard1988 Professional Programmer Join Date: Oct 2005 Location: Chitown Posts: 422 Rep Power: 4	One of our members has a good explanation . http://www.eternallyconfuzzled.com/a..._art_rand.aspx __________________

Mar 26th, 2007, 2:08 AM	#7
Game_Ender Professional Programmer Join Date: May 2006 Location: Maryland, USA Posts: 306 Rep Power: 3	DaWei's point is that it is possible with the right hardware and libraries to get true random number in C. So just because we are talking about C and random numbers does not mean they are always pseudo-random. __________________ Robotics @ Maryland AUV Team - Software Lead

Mar 26th, 2007, 2:10 AM	#8
DaWei Resident Grouch Join Date: Jun 2005 Posts: 6,453 Rep Power: 10	You are apparently referring to the PRNG commonly found in the run-time library. This does not preclude one from using C with true random numbers. I'm of the opinion that disseminating flakey information, whether by loose semantics or otherwise, is not a good idea. Again, I'd suggest you follow the earlier link for additional considerations involved in seeding the PRNG with the system's time. __________________ Abstraction doesn't make it impossible to write bad code; it makes it possible to write superior code. Contributor's Corner: Grumpy on C++ Exceptions DaWei on Pointers

Mar 26th, 2007, 1:47 AM	#4
Fall Back Son Professional Programmer Join Date: Oct 2006 Posts: 311 Rep Power: 3	short version: Random numbers aren't really random. They are either "seeded" by a number or by the system time. If seeded by a number, they use calculations to come up with the "random" numbers. If seeded by the system time, they use that to come up with the random numbers. The system time method is better in some cases because it is closer to true randomness.

Mar 26th, 2007, 2:04 AM	#6
Fall Back Son Professional Programmer Join Date: Oct 2006 Posts: 311 Rep Power: 3	I was clearly referring to 'random numbers' in C programming. Does that clear it up, or am I still incorrect?

Mar 26th, 2007, 2:16 AM	#9
Fall Back Son Professional Programmer Join Date: Oct 2006 Posts: 311 Rep Power: 3	Can you show me such a code? Not saying it doesn't exist, but it seems that any code would not be truly random. "I'm of the opinion that disseminating flakey information, whether by loose semantics or otherwise, is not a good idea." edit: DaWei, I'm all for correct grammar as well, but couldn't you just tell me to shut the hell up if I don't know what I'm talking about? Also: I already read everything in the original link. I'm still of the opinion that the numbers you say are "truly random" are only more significantly random than others. While I understand the article's point about distribution, a perfect distribution does not necessarily indicate true randomness. Is that incorrect?

Mar 26th, 2007, 9:40 AM	#10
DaWei Resident Grouch Join Date: Jun 2005 Posts: 6,453 Rep Power: 10	I didn't mention grammar. The statement was made that C can use processes that generate truly random numbers. No one mentioned that this will happen with any PRNG, regardless of it's distribution. My view on misinformation remains the same; my expression seemed preferable, at the time of that post, to telling you to just shut the fuck up, read the posts, and maybe resort to Google. __________________ Abstraction doesn't make it impossible to write bad code; it makes it possible to write superior code. Contributor's Corner: Grumpy on C++ Exceptions DaWei on Pointers

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Median/Mode in arrays? {Need help}	Java\|Tera	Java	27	Nov 29th, 2005 11:50 AM
FiveDigit + RandomeNumber Game.	TecBrain	Java	0	Nov 18th, 2005 3:53 PM
Random Number	coldDeath	Python	1	Sep 11th, 2005 9:17 AM
Random Number & Average Problem	Hadrurus	Java	6	Aug 15th, 2005 2:08 PM
non repeating random number generation	gencor45	C#	2	Feb 9th, 2005 1:11 AM

Programming Forums > Application Development > C

Can anyone explain Random number generation