Hello!
That's not a bug, this works as intended. In python all numbers are "two complement". Many other programming languages behaves the same way (JavaScript, C, Rust and others).
For your case you should use operator.neg:

>>> import operator
>>> operator.neg(60)
-60

Changing it would be break backward compatibility with a lot of python code.

Hi @mohamadiarch , I think you confuse couple of things:

we just convert ones and zeros to each other and nothing else, and that's a one's complement number.

Under the hood, signed integers are represented in 2’s complement format. When you perform ~, you don’t change signed integer representation from 2’s to 1’s complement. You just flip the bits. Maybe the fact that ~ is also called ‘complement’ confused you. But anyway, here is further explanation:

Negation of signed integer x is defined as:
  $-x = flipbits(x) + 1$
Since ~ operator flips the bits (makes complement) of the integer we can express it as:
  $-x = \sim x + 1$
Which means that:
  $\sim x = -x - 1$

This is why ~60 == -61, which is correct.

@Eclips4

Changing it would be break backward compatibility with a lot of python code.

Not only that but whole integer system in general :)

@WolframAlph Can you break it down with an example 60? I think if we use flipbits(x) + 1 for 00111100 (60 in decimal) we should have 11000100 and is this means "-61"?

@mohamadiarch, 11000100 binary is -60 decimal in 2’s complement format. flipbits(x) + 1 negates x, remember?
So:

• flipbits(60) + 1 == -60.
• flipbits(60) == -61

Makes sense? Flipping bits of some integer and negating integer is not the same. Please, read through my explanation carefully.

@WolframAlph I have understood your formula. But what I can't figure out is the method of going from binary to decimal. You use the formula for the final answer and get -61, but I want to read it without it. You said: You just flip the bits and I think this is an introduction to one's complement. If I flip all the bits of 00111100 I get 11000011. Now if I want to read this number I have two choices. Reading from 1's complement or reading from 2's complement. For the first one, I get -60 and for the second one, I get -61. But if we use the formula we get -61.

@mohamadiarch

Now if I want to read this number I have two choices

Correct.

Reading from 1's complement or reading from 2's complement

Wrong. Have you read what I wrote earlier?

Under the hood, signed integers are represented in 2’s complement format.

So yes, you can interpret 11000011 binary number in 2 ways. But those are:

1. Interpret as 8bit signed integer - -61
2. Interpret as 8bit unsigned integer - 195

Interpretation depends on context. In this scenario context is that x = 60 is positive and therefore ~x will produce negative number, because if x >= 0 then ~x = -x - 1 produces negative result. This is why Python tells you it is -61 instead of 195. In summary: computer sees signed integers in 2’s complement and you cannot expect to get correct results by interpreting as 1’s complement. My examples were for 2’s complement.

invert operator ~: "The bitwise inversion of x is defined as -(x+1). " @mohamadiarch In the future, please consult the docs. A place to discuss such details is https://discuss.python.org/c/users/7.

@WolframAlph I made a mistake. I'm sorry, I shouldn't have said that in this place. Thank you for your explanation. Your explanation was perfect.