On Mon, Apr 30, 2001 at 02:13:08PM -0400, Larry Hall (RFK Partners, Inc) wrote:
> At 01:37 PM 4/30/2001, you wrote:
> >void main()
At the risk of being accused of pendantry, main() should be declared as
int.
> >{
> > int input1, i;
> >
> > srand(time(NULL));
> >
> > for (i=0; i<20 ; i++)
> > {
> > input1 = rand() * 10 / (RAND_MAX + 1);
Here's your problem. You're doing integer math. 10/(RAND_MAX + 1) is
going to be 0. Think about why . . you're dividing a small integer by a
rather large integer.
Moreover, RAND_MAX is usually an alias for INT_MAX. Therefore RAND_MAX
+ 1 is an overflow condition.
Moreover, you don't want the +1 in the denominator. You want rand()'s
proportion of RAND_MAX--which (should be) a uniform deviate, multiplied
by 10, to bring its range into 0 < 9, and THEN add 1, which will (after
truncation) give you numbers between 1 and 10.
I think what you really want is something like this:
input1 = (double)rand() * 10.0/((double)RAND_MAX) + 1.0;
> > printf("random numbers %i\n",input1);
> >
> > }
> >}
> Try the latest snapshot at www.cygwin.com.
The OP`s problem was semantic. He was specifying integer math, but
expecting results of double math. According to integer math, his result
is correct . . though not what he intended.
Charles
--
Charles Krug, Jr.
Applications Engineer
Pentek Corp
1 Park Way
Upper Saddle River, NJ 07458
--
Want to unsubscribe from this list?
Check out: http://cygwin.com/ml/#unsubscribe-simple
