round

[Peter Otten]

ast wrote:

> Hi
> 
> round is supposed to provide an integer when
> called without any precision argument.
> 
> here is the doc:
> 
>  >>> help(round)
> 
> round(number[, ndigits]) -> number
> 
> Round a number to a given precision in decimal digits (default 0 digits).
> This returns an int when called with one argument, otherwise the
> same type as the number

That's not the complete story. Quoting 
https://docs.python.org/dev/library/functions.html#round

"""
For a general Python object number, round delegates to number.__round__.
"""

Bogus example to make the point:

>>> class A:
...     def __round__(self): return "whatever"
... 
>>> round(A())
'whatever'

> but in some circumstances it provides a float
> 
> import numpy as np
> 
> M = np.array([[0, 9],[2, 7]], dtype=int)
> np.linalg.det(M)
> -18.000000000000004
> round(np.linalg.det(M))
> -18.0 # i was expecting an integer -18, not a float
> 
> # same problem with np.round
> np.round(np.linalg.det(M))
> -18.0

>>> M = np.array([[0, 9],[2, 7]], dtype=int)
>>> type(np.linalg.det(M))
<class 'numpy.float64'>

So numpy.linalg.det() returns a custom type float64 which maps round() to 
float64:

>>> round(np.float64(1.23))
1.0
>>> type(_)
<class 'numpy.float64'>