fit_one_of function to avoid redundant parameters

[richardh]

Could put some sort of function in model.py to identify which parameters should not all fit simultaneously, then flag to user as an issue.

e.g. for a simple particle:

fit_one_of('scale', 'sld', 'sld_solvent')

e.g. for a core-shell particle, think might need:

fit_one_of('scale', 'sld_core'.AND.'sld_shell'.AND.'sld_solvent')

There may be a better way to express this, or a clever way to to this automatically, certainly from the derivatives, or the correlation matrix. Also does not work always, e.g. if sld_shell is fixed at same as sld_solvent then the core/shell system beomes a simple particle, but could add extra logic for this???