Advertisement
I'm writing a few short programs based on calculating factors in electronics.
As such, I often have to multiply and divide some small numbers, and the round-off is of such a magnitude that it renders the calculations useless.
For instance, with this fragment of code:
#include <iostream>
using namespace std;
int main () {
double test;
test = ( 1 / 10000) + ( 1 / 200000);
cout >> test;
return 0;
}
The output will give me zero.
As such, I often have to multiply and divide some small numbers, and the round-off is of such a magnitude that it renders the calculations useless.
For instance, with this fragment of code:
#include <iostream>
using namespace std;
int main () {
double test;
test = ( 1 / 10000) + ( 1 / 200000);
cout >> test;
return 0;
}
The output will give me zero.
posted by:
|
Unsubscribed |
Advertisement
Advertisement
-
Re: Floating point woes
Mon, September 4, 2006 - 10:39 AM1, 10000, and 200000 are all integers, so the math is done using integers. Then the result (0) is converted to a double when assigned to 'test'
It needs to look like this:
test = (1.0 / 10000.0) + (1.0 / 200000.0);
If you are using variables, make sure they are cast to doubles, for example:
int i = 10000;
test = 1.0 / (double)i;
-
-
Unsu...
Re: Floating point woes
Tue, September 5, 2006 - 12:26 PMAh-ha! Beautiful! Thank you very much!
-
