Improve heuristics for chosing default number of scattering angles (-ntheta) #100
Description
Currently this heurestics is based on number of dipoles. This becomes
especially inadequate when refining discretization of the fixed particle.
Much better alternative is to base it on size parameter that is also readily
available in the code. Because size parameter determines the typical angular
scale for variation of light scattering patterns (from diffraction-like
arguments).
Original issue reported on code.google.com by yurkin on 8 Jun 2010 at 4:01