Piecewise bug with scalar domains and more than 2 conditions

(Commented in #4792)
There, seems to be an issue with scalar domains and three or more conditions:

>>> np.piecewise(3, [True, False, False],[4,2,0])
ValueError: function list and condition list must be the same

Funny enough it works fine with two conditions (this is what has been tested):

>>> np.piecewise(3, [True, False],[4,2])
array(4)

The three conditions version works fine with select:

>>> np.select([True, False, False],[4,2,0])
array(4)

Tested on python 3.4 with numpy 1.9.0 and python 2.7.6 with numpy 1.8.2

Provide feedback