[argerlt]

Description of the change

Solves #501. @hakonanes and/or @viljarjf , let me know what you think of this approach. if we like it, I will address the failing unit tests add some new ones, and include documentation.

Main features of this PR:

• The Miller dunder methods __neg__, __add__, and __sub__ no longer discard phase information or coordinate preference
• Warnings have been added using the new logger feature that alert users when symmetry or coordinate frame mismatches occur.
• Simplification of Vector3d._radd__, Vector3d._rsub__ , Vector3d._sub__ and Vector3d.__rmul__ to avoid repeating code and allow inheriting phase information from a Millerclass.

Progress of the PR

• Docstrings for all functions
• Unit tests with pytest for all lines
• Clean code style by running black via pre-commit

Minimal example of the bug fix or new feature

from orix.vector import Miller, Vector3d
from orix.crystal_map import Phase

p1 = Phase(point_group="m-3m")
p2 = Phase(point_group="6mm")

m_xyz = Miller(xyz=[1, 2, 1], phase=p1)
m_hkl = Miller(hkl=[1, 2, 1], phase=p1)
m_hex_hkl = Miller(hkl=[1, 2, 1], phase=p2)

v = Vector3d([1,2,3])

# Negation copies symmetry and coordinate format
print(m_xyz)
print(-m_xyz)

print(m_hkl)
print(-m_hkl)

# Adding mismatched crystal vectors throws appropriate warnings
print(m_xyz + m_hkl)
print(m_hkl + m_hex_hkl)
print(m_hex_hkl + m_xyz)

# combining crystal and real space vectors always returns a crystal vector,
# but with a warning.
print(m_xyz + v)
print(v + m_xyz)
print(m_xyz - v)
print(v - m_xyz)

# additionally, sample vectors are assumed to have the same coordinate format as
# the crystal vector they are being combined with.
print(m_hkl + v)
print(v + m_hkl)
print(m_hkl - v)
print(v - m_hkl)

For reviewers

• The PR title is short, concise, and will make sense 1 year later.
• New functions are imported in corresponding __init__.py.
• New features, API changes, and deprecations are mentioned in the unreleased
  section in CHANGELOG.rst.
• Contributor(s) are listed correctly in __credits__ in orix/__init__.py and in
  .zenodo.json.

[argerlt]

pre-commit.ci autofix

[argerlt]

Addendum: Maybe we don't want to get into it now, but the latest commit would eliminate the need for #499 and #500 as well.

IMO, the related thread on diffsims was about if you should or shouldn't do the given example calculation, which is an important question for Diffsims, but handling it on the ORIX side is a far simpler single-line fix. @viljarjf 's original solution is basically what I did here to get Vector+Miller to always return a Miller, so it seems like a great solution now, whereas I believe #499 breaks some newer orix features.

If anyone disagrees, let me know and I'll revert it.

[argerlt]

classes work as expected now, just need to write unit tests and examples.