FloatingPointError: encountered floating-point exception

FloatingPointError: encountered floating-point exception

$ python -c "import numpy as np; np.seterr(all='raise'); np.array([1.0]) / 0"
Traceback (most recent call last):
  File "<string>", line 1, in <module>
FloatingPointError: divide by zero encountered in true_divide

Why this happens

NumPy can be configured to raise exceptions on floating-point issues (divide-by-zero, invalid operations, overflow). When such an operation occurs, FloatingPointError is raised.

Fix

Handle or avoid illegal operations, or set error handling back to warnings using np.seterr. Use masked arrays or check denominators before dividing.

Wrong code

import numpy as np
np.seterr(all='raise')
print(np.array([1.0]) / 0)

Fixed code

import numpy as np
np.seterr(all='warn')
den = np.array([0.0])
num = np.array([1.0])
safe = np.where(den == 0, np.nan, num/den)
print(safe)