Summary

np.exp raises AttributeError: 'int' object has no attribute 'exp' when called with an integer greater than 2**64-1.

Expected behavior

Return np.inf as for other large integer values (e.g. for 2**63).

Assumption

Such integers are treated internally as object and hence numpy tried to call the exp-function on the instance (similar to np.exp(np.asarray([2], dtype=object))). However this is completely unexpected since in Python 3 all integers have the same type (nonetheless the must be stored differently in memory).

System

$ uname -a
Linux MyPC 4.4.0-128-generic #154-Ubuntu SMP Fri May 25 14:15:18 UTC 2018 x86_64 x86_64 x86_64 GNU/Linux
$ python
Python 3.6.2 (default, Jul 17 2017, 23:14:31) 
[GCC 5.4.0 20160609] on linux

Code Example

>>> from numpy import exp
>>> exp(2**64)  # Would expect np.inf as return value.
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: 'int' object has no attribute 'exp'
>>> exp(2**63)  # Here it does.
__main__:1: RuntimeWarning: overflow encountered in exp
inf
>>> type(2**63) is type(2**64)
True