Fixed default argument for refr. index in Steps()

[ldoyle]

Fixes the problem that the default arg. for refractive index was a scalar. The new behaviour is:

• If no argument for refr is supplied, it defaults to 1.0 (as before)
• If no argument is supplied or a (complex) scalar is supplied, like 1.33 or 1.0-1j, this number is expanded to a 2D grid of correct size with all elements having the same value (homogeneous medium)

NB: As mentioned in the manual (but forgotten in the meantime by me), the change of optical path length by a real refractive index is not fully taken into account, so a homogeneous medium with refr=1.33 yields the same result as refr=1.0 and therefore is only interesting with varying refractive index. But of course this fix still makes sense for complex refr. index.

[FredvanGoor]

solved "refractive index scalar" bug in Steps command