orix 0.12.0 is a minor release of orix, an open-source Python library for handling orientations, rotations and crystal symmetry.

See below, the changelog or the GitHub changelog for all updates from the previous release.

Added

Vector3d.from_path_ends() class method to get vectors between two vectors.
Convenience function plot.format_labels() to get nicely formatted vector labels to use when plotting vectors.
Two offsets in the stereographic coordinates (X, Y) can be given to StereographicPlot.text() to offset text coordinates.
Explicit support for Python 3.11.
Creating quaternions from neo-eulerian vectors via new class methods from_rodrigues() and from_homochoric(), replacing the now deprecated from_neo_euler(). from_rodrigues() accepts an angle parameter to allow passing Rodrigues-Frank vectors.
Creating neo-eulerian vectors from quaternions via new methods to_axes_angles(), to_rodrigues() and to_homochoric(). Rodrigues-Frank vectors can be returned from to_rodrigues() by passing frank=True.
inv() method for Quaternion, Rotation, Orientation, and Misorientation. For the three first, its behavior is identical to the inversion operator ~. For misorientations, it inverts the direction of the transformation. Convenient for chaining operations.
The random() methods of Orientation and Misorientation now accept symmetry. A random() method is also added to Vector3d and Miller, the latter accepting a phase.
Function orix.sampling.get_sample_reduced_fundamental() for sampling rotations that rotate the Z-vector (0, 0, 1) onto the fundamental sector of the Laue group of a given Symmetry.

Changed

The convention parameter in from_euler() and to_euler() will be removed in the next minor release, 0.13, instead of release 1.0 as previously stated.
Allow passing a tuple of integers to reshape() methods of 3D objects.
random() methods no longer accept a list as a valid shape: pass a tuple instead.
Increase minimal version of Matplotlib to >= 3.5.

Creating quaternions from neo-eulerian vectors via from_neo_euler() is deprecated and will be removed in v0.13. Use the existing from_axes_angles() and the new from_rodrigues() and from_homochoric() instead.

Transparency of polar stereographic grid lines can now be controlled by Matplotlib's grid.alpha, just like the azimuth grid lines.
Previously, Phase did not adjust atom positions when forcing Phase.structure.lattice.base to use the crystal axes alignment e1 || a, e3 || c*. This is now fixed.