Håkon Wiik Ånes

I like the interpretation of `o2 * ~o1` as a matrix multiplication because an orientation (in orix) is a transformation from the sample to the crystal: first we invert `o1`...

I think your rules look reasonable, apart from rule 3 which I don't understand: Will a resulting misorientation's `_symmetry` attribute be a tuple of 2-tuples of two symmetries, as in...

Thanks for mentioning this, @din14970. If you have the time, could you show how to create this Al6Mn structure with ASE, which I'm here doing with diffpy.structure: ```python >>> from...

> Ah, the various opening and closing got me confused (thought this was in diffsims) Sorry for that! > I'm more receptive to such a refactor in orix than I...

Thanks, initialization is simpler with ASE than with diffpy.structure, it seems. I couldn't immediately find in the docs or via inspecting the `crystal` instance how to compute reciprocal/direct lattice vector...

> Hmmm, seems indeed not as native to do such things with ASE. I think the main difference is that ASE started from the principle of arbitrary atomic models, not...

> I think it becomes clear that ASE is forcing a parallel x-vector whereas diffpy is forcing a parallel z-vector. This ties in with our discussion on crystal axes alignment...

Right, I think I'm the one who only discusses this in context of orix, as I think @din14970's original comment was in terms of pyxem packages in general. It's good...

This is good to know, and a note stating this should perhaps be added to the [crystal reference frame user guide](https://orix.readthedocs.io/en/stable/crystal_reference_frame.html)? ASE specifies their crystal axes alignment in [ase.geometry.cellpar_to_cell()](https://wiki.fysik.dtu.dk/ase/ase/geometry.html#ase.geometry.cellpar_to_cell). This...

Good question, and yes the use case is when you want to only update some of the rotations by using a boolean mask into `CrystalMap.rotations`.