ChemoDynamicalEvolution

My approach would be to avoid a configuration file and just have a general ChemoDynamicalEvolution class that can be supplied with instances of things like a TimeDelayDistribution class, a TypeIISNeYields class, TypeIaSNeYields class, AGBWindYields class, an IMF class, etc. I'm not entirely sure about how to best allow gas flows to be represented. For stellar migration, we would also like to track stellar (x,v) or at least actions or such. We can then use diffusion equations to describe the evolution of stellar orbits (and we would also have to specify the dynamical state of star forming gas). I typically like to keep a python-only core and interface, to allow for easy installation and experimentation with adding new features.

I'm a little worried that creating a full-featured model would be a larger project then either of us wants to undertake at this point (seems perfect for a student... like Brett I guess!). Perhaps we should focus at first on setting up a simple model with, say, independent zones for which gas in/outflow, the SFH, the metallicity of the inflowing gas, etc. can be specified. Radial migration could be allowed in post-processing (without tracking returned metals from type Ia SNe going off somewhere very different from where they formed). The inputs for these simple models could then be taken from the simulations.