Max_rand

This is a discussion on Max_rand within the C++ Programming forums, part of the General Programming Boards category; Hey guys, I'll be googling and such and will probably find the answer before someone can reply, but in case ...

  1. #1
    Registered User
    Join Date
    Mar 2006
    Posts
    12

    Max_rand

    Hey guys, I'll be googling and such and will probably find the answer before someone can reply, but in case this problem is trickier than it seems, here's the problem I am having.

    Basically, I never needed to have a random number that was bigger than 0x7fff (a little over 32000 in decimal) until now. But I want to be able to generate a random number that is say, a number between 0 and 1000000. How can I do this, and still have the number be truly random, i.e. each number between 0 and 1000000 is just as likely?

    edit: I think I just figured it out just by thinking about it

    the answer is rand() * MAX_RAND + rand()
    Last edited by psyadam; 10-25-2007 at 04:41 PM.

  2. #2
    Hardware Engineer
    Join Date
    Sep 2001
    Posts
    1,398
    FYI - It's actually RAND_MAX, rather than MAX_RAND.

    the answer is
    Code:
     rand() * MAX_RAND + rand()
    Hmmmm.... I don't think that will give you an even-uniform distribution.. I think that's going to favor big numbers, but I'd better leave that one for the experts.

  3. #3
    Registered User
    Join Date
    Apr 2006
    Posts
    2,021
    I think you want something like this:
    rand() << lg2((long)RAND_MAX+1) + rand()
    Which simplifies to this:
    (long)rand()<<32 | (short)rand()
    It is too clear and so it is hard to see.
    A dunce once searched for fire with a lighted lantern.
    Had he known what fire was,
    He could have cooked his rice much sooner.

  4. #4
    C++まいる!Cをこわせ! Elysia's Avatar
    Join Date
    Oct 2007
    Posts
    22,450
    Huh... What a nice way to generate a high number that was never intended to be generated by the function. I like it!

  5. #5
    Algorithm Dissector iMalc's Avatar
    Join Date
    Dec 2005
    Location
    New Zealand
    Posts
    6,295
    Quote Originally Posted by King Mir View Post
    I think you want something like this:
    rand() << lg2((long)RAND_MAX+1) + rand()
    Which simplifies to this:
    (long)rand()<<32 | (short)rand()
    That shift can't be supposed to be by 32 or else he would already be able to generate big enough random numbers. You must mean 16.
    My homepage
    Advice: Take only as directed - If symptoms persist, please see your debugger

    Linus Torvalds: "But it clearly is the only right way. The fact that everybody else does it some other way only means that they are wrong"

  6. #6
    Registered User
    Join Date
    Apr 2006
    Posts
    2,021
    Er, yeah.
    It is too clear and so it is hard to see.
    A dunce once searched for fire with a lighted lantern.
    Had he known what fire was,
    He could have cooked his rice much sooner.

  7. #7
    Registered User
    Join Date
    Mar 2006
    Posts
    12
    I wrote some code to test the uniformity of the distribution of my random number generator and it appears to be uniform. My first idea, btw, was to use rand() * rand() + rand(), which does not give you a uniform distrubution

    It gives you a distribution like this:

    0: 329470
    1: 192222
    2: 139096
    3: 105084
    4: 80234
    5: 59911
    6: 43583
    7: 28865
    8: 16319
    9: 5216
    Press any key to continue

    where the numbers 0-9 represent 10 equal ranges for the LargeRand() function

    (edit: i.e. if SUPER_RAND_MAX was 1,000,000, 0 would represent the number of numbers in the range 0 to 100,000, 1 would represent the number of numbers in the range 100,000 to 200,000 etc)
    Last edited by psyadam; 10-27-2007 at 12:03 AM.

Popular pages Recent additions subscribe to a feed

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21