Allow resampling of SFH and metallicity functions

[stephenmwilkins]

At the moment we have a function that resamples a SFZH array. However, it would be nice to also have a function that sampled a SFH and metallicity function instead.

This could be achieved either in the new function or through a e.g. sample method included in the SFH/Z classes themselves (or both, i.e. with a wrapper that seamlessly produces a new particle.stars object).

[WillJRoper]

I'd say choose a lane, and do one of the other. A method is preferable IMO. All these redundant ways of doing things seem great until you realise you have an incomprehensible mess to document which just confuses a user.