In the process of implementing a universal kspacing definition, I discovered that the various DFT calculators do all sorts of different things when they encounter non-periodic systems. It would be nice if all these behaviors were rationalized and made consistent, although it's possible that the input params are so different that this won't actually achieve much. But I think it's worth a review.
This may not be as bad as I had initially thought. I'm working on a PR that will address this as well as #258 and the universal kspacing heuristic that's mentioned in #255
In the process of implementing a universal kspacing definition, I discovered that the various DFT calculators do all sorts of different things when they encounter non-periodic systems. It would be nice if all these behaviors were rationalized and made consistent, although it's possible that the input params are so different that this won't actually achieve much. But I think it's worth a review.