You had 0/0. Maybe you want an exception to be raised, but floating point standard has that "NaN" is the result.
Numpy has an optional warning message, if I remember, but I don't know if there is a performance penalty for checking every division floating point computation.
