Floating point subtraction adds weird trailing value

Asked Feb 05 '17 at 19:06

Active Feb 05 '17 at 19:06

Viewed 40 times

I'm sure there is an answer out there but I just can't find it.

Why does python add 0.000000000000000009 to this floating point subtraction?

Python 2.7.12 (default, Oct 10 2016, 21:41:23)
[GCC 4.2.1 Compatible Apple LLVM 8.0.0 (clang-800.0.38)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> 1.01-1
0.010000000000000009
>>> 1.01-1.0
0.010000000000000009
>>> (1 + 0.01)-1
0.010000000000000009
>>>

asked Feb 05 '17 at 19:06

Realistic

1,038
1
10
20

2

Simply because floating points cannot represent `0.01` exactly. – Willem Van Onsem Feb 05 '17 at 19:06
Thanks @vaultah that was exactly what I was looking for – Realistic Feb 05 '17 at 19:09
1

Actually the subtraction is exact. It's `1.01` that's already wrong. You can see it with `'%.60f' % 1.01`. And if you compare it with `'%.60f' % (1.01 - 1)`, you'll see that the actual difference is exactly 1. – Stefan Pochmann Feb 05 '17 at 19:14
1

@WillemVanOnsem I don't see what this has to do with `0.01`. That neither appears here as literal nor as result (`1.01 - 1 == 0.01` is `False`). – Stefan Pochmann Feb 05 '17 at 19:32

Floating point subtraction adds weird trailing value

0 Answers0