Question

I’m writing a code that performs a random scan over a set of 4 numbers. I’d like to scan 10000 points (millions later). I just learned about rand(), so here’s the relevant part:

int numPoints = 10000;
double A,B,C,D;
 for (i=0; i<=numPoints1;i++) {

                srand ( time(NULL) );
                A = rand() % 500 + 100;
                B = rand() % 500 + 100;
                C = rand() % 100 - 100;        
                D = rand() % 5 + 2.5; 

Then these four variable are fed into a function (A,B,C,D).
The code performs some checks and calculations inside the loop.

 }

However, I noticed in the output file that many times, the same set of A,B,C,D is picked.

Q: Any suggestions as to how can I improve the situation?

Answer 1

You should move the call to srand out of the loop. Like this:

srand(time(NULL));
for (i = 0; i <= numPoints1; i++)
{
   ...
}

What’s happening is that you’re re-initializing the random number generator with each iteration. As I recall, time returns the time as a number of seconds. So time(NULL) will only change once per second, meaning that you’ll be seeding the random number with the same seed multiple times.

That said, this won’t guarantee that a set of numbers isn’t repeated. It will, however, make duplicates much less likely.