Casting big NSInteger to float issue

Question

1) define big NSInteger
2) cast to float
3) log float
4) turns into slightly smaller number every time???

NSInteger val = 485935336;
float val_float = (float)val;
NSLog(@"%f", val_float); //logs 485935328, not 485935336

---

Any clue why?

Answer 1

Any idea why? Of course. float has about six digits of precision. double has 15 digits precision, which is why you should always use double unless you can give a good reason why not. You will get the same problem, but only when the numbers are a lot bigger.