Christoph Ortner

> of the space the atoms are in ... I understand the intention fine. I was just trying to point out that the interpretation is not quite as clear-cut as...

I think that means we keep what we have now, document clearly what D means and then close the issue?

To my ears this is ambiguous for the reasons I explained above. "Space " for a mathematician and actually many physicists as well is a much broader concept.

@mfherbst Good point. We will store training data in structures. Not sure how else to do this. Can there be a data dict which is excluded in the == ?

Could also ≈ be added in a reasonable way?

I'd love to have this PR, I'm current hand-writing equality and ≈ tests.

It feels all these methods to manipulate systems could become part of `AtomsBuilder`

I'm planning to move over the rest of JuLIP functinality to AtomsBuilder over the coming week. If you file an issue there and describe how you think `wrap` should work...

also see #97 and #99 -- apologies for duplicating discussions.

Very interesting option, but an implementation detail and easy enough to provide. In the interface update for 0.4.x we can just allow arbitrary indices rather than only `Integer` (needs a...