Rand Implementation

Question

I would like to go through how rand() and srand() functions are implemented and would like to tweak the code to modify it to my requirements. Where can i find the source cod

Answer 1

The glibc one (used by gcc) is the simple formula:

x = 1103515245 * x + 12345

wrapping around at 2³², as shown here. You can just set x as the seed then keep calling a function to evaluate that expression (and update the seed).

But you should be aware the linear congruential generators like this are considered adequate but not ideal.

While the only ideal random number generator would be perfectly random, the Mersenne Twister probably comes closer.

Answer 2

It takes a seed as in input argument, usually like follows:-

double result = srand(time(NULL));

and returns a random number that adheres to the probability and hence expected number of occurrences.

from CodeGuru forums:-

void __cdecl srand (unsigned int seed)
{
    #ifdef _MT
        _getptd()->_holdrand = (unsigned long)seed;
    #else /* _MT */
        holdrand = (long)seed;
    #endif /* _MT */
}

int __cdecl rand (void)
{
   #ifdef _MT
    _ptiddata ptd = _getptd();
    return( ((ptd->_holdrand = ptd->_holdrand * 214013L + 2531011L) >> 16) &
    0x7fff );
   #else /* _MT */
    return(((holdrand = holdrand * 214013L + 2531011L) >> 16) & 0x7fff);
   #endif /* _MT */
}

Hope this helps.

Answer 3

rand and srand are usually implemented as a simple LCG, you can easily write your own (it's few lines of code) without looking for the sources of rand and srand. Notice that, if you need random numbers for "serious" purposes (e.g. cryptography), there are much better RNGs than LCG.

By the way, the C standard itself includes a sample implementation of rand and srand:

static unsigned long int next = 1;

int rand(void) // RAND_MAX assumed to be 32767
{
    next = next * 1103515245 + 12345;
    return (unsigned int)(next/65536) % 32768;
}

void srand(unsigned int seed)
{
    next = seed;
}