Interpolation for 1D Slices

[jfields7]

The interpolator in PyCompOSE uses cubic splines when computing things like beta-equilibrium slices for initial data, but our evolution codes generally use linear interpolation. This can create inconsistencies between the initial data and the actual evolution which present themselves as small temperature violations.

Should we consider having an option for doing linear interpolation, perhaps making it the default? I personally tend to be opposed to doing high-order interpolation with tables as is because it's hard to enforce monotonicity where it's needed, but I don't know if there are other strong opinions about this.

[dradice]

I think it would be a good idea to add an optional field for the interpolation to the relevant functions. We could even set the default to linear interpolation, so that we get consistent initial data. The best thing would be to have thermodynamically consistent interpolation of the EOS in the code, but that is not happening anytime soon.