Odd pow Behavior

[ichabod801]

I was doing some unit testing and came across something odd:

>>> -2 ** 4
-16
>>> pow(-2, 4)
16
>>> import operator
>>> operator.pow(-2, 4)
16
>>> import math
>>> math.pow(-2, 4)
16.0

All of the documentation says that the various pow functions are equivalent to x ** y. However, the function call changes the order of operations. The negation is carried out when the number is evaluated as a parameter, instead of after the ** operator. So the pow functions are really calculating (x) ** (y), which is slightly different.

[micseydel]

That doesn't seem odd to me. It's just operator precedence. I might agree though that the exact docs should be updated.

I was more surprised that math.pow has a float result for what could have been an int, whereas the previous results all stayed ints.