CODEWITHSUNDEEP
pythonnumpy
TheShin
TheShin

Reputation: 77

My function calculating a negative number even when it's not supposed to

I have a NumPy array [545, 32] and I pass a function to it:

def obj_func(array):

    x = array[0]
    y = array[1]
    obj_func = (x**2 + y - 11)**2 + (x + y**2 - 7)**2
    return obj_func

The formula should look something like

(545^2 + 32 - 11)^2 + (545 + 32^2 - 7)^2

The answer that I got when I passed it through my function is -1955547256, but the answer should be about 8.82387x10^10.

It probably has something to do with manipulating big numbers, since the smaller values I calculated through my function had the correct answer. But why did it return a negative number, etc.? I know for sure this function should not return any negative numbers.

Upvotes: 0

Views: 84

Answers (1)

mathfux
mathfux

Reputation: 5949

You might like whether to change type of variables in you array or call obj_func on list instead:

obj_func(np.array([545, 32]))
>>> -1955547256

obj_func([545, 32])
>>> 88238765960

obj_func(np.array([545, 32], dtype=np.int64))
>>> 88238765960

This effect happens when data of array exceeds maximum machine limit for array.dtype = np.int32

It can be inspected like so:

np.iinfo(np.int32)
>>> iinfo(min=-2147483648, max=2147483647, dtype=int32)
np.iinfo(np.int64)
>>> iinfo(min=-9223372036854775808, max=9223372036854775807, dtype=int64)
np.array([2147483647]) + 1
>>> array([-2147483648]) # shifts to minimum machine limit of dtype=int32

Upvotes: 2

Related Questions