In Python 3.8, when I run: -math.inf + math.inf ==> I am getting "nan".
Should the result be: 0? Or what is the rational for Python to produce "nan" for this addition?
Asked
Active
Viewed 128 times
-1
-
5I'd look at the IEEE math rules for that operation. – Mark Ransom Mar 24 '22 at 00:54
-
2Here link to the [IEEE standard on wikipedia](https://en.wikipedia.org/wiki/IEEE_754#2019). You might also be intrested in the following [thread](https://stackoverflow.com/questions/36151158/how-are-nan-and-infinity-of-a-float-or-double-stored-in-memory) explaining the bit representation of `+inf`, `-inf` and `NaN` and providing further links. – Mushroomator Mar 24 '22 at 01:14
1 Answers
1
In mathematical terms Infinity is not a definable number. You're thinking of it as "a really big number that has a value". It doesn't. (Infinity * Infinity) is still Infinity. Don't try to mentally assign a real value to this. The "nan" value makes sense as any math with Infinity (or math.inf in this case) really makes no sense.
stdunbar
- 16,263
- 11
- 31
- 53
-
-
I've another question regarding math.inf in Python: https://stackoverflow.com/questions/71595929/in-python-how-to-override-the-arithmetic-operator-to-produce-1-0-math – charlesqwu Mar 24 '22 at 01:15
-
Are you saying [this](https://en.m.wikipedia.org/wiki/Extended_real_number_line) can't be done? – Kelly Bundy Mar 24 '22 at 01:51