Confusing documentation for remainder and fmod

[callegar]

Current numpy documentation for fmod says that "This is the NumPy implementation of the Python modulo operator %"

But this is confusing, since fmod and % give different results

fmod(-3,2) -> -1
-3 % 2 -> 1

Furthermore, the np.remainder documentation makes no mention of fmod, nor of sign conventions. This is bad, since users coming from Matlab may use remainder to translate the Matlab 'rem', while the two things are not the same.