v2.1.8

Release 2.1.8

• Added explicit testing with python 3.10.
• Cleaned up unused imports.
• Updated some modules which were using deprecated code.

v2.1.7

Release 2.1.7

• Fixed an issue where :func:gaussian would trip on a full-width at half-maximum of 0.
• Fixed an issue where the first stage of the dual-tree complex wavelet transform was not shifted properly (#36).

v2.1.5

Release 2.1.5

• Releases are now automatically performed using Github Actions
• It is now possible to install all the dependencies required to use :func:diffshow using the following installation option: pip install scikit-ued[diffshow].

Release 2.1.4

This release only brings one minor change:

• Increased the reliability of bragg_peaks to distinguish between noise and Bragg peaks.

Release 2.1.3

This release brings one new feature:

• Added the function :func:bragg_peaks to determine the location of single-crystal diffraction peaks in an image.

and one maintenance tweak:

• Fixed deprecation warnings regarding NumPy's dtypes.

Release 2.1.2

• Improved autocenter for diffraction patterns with large Ewald sphere walkoff.
• diffread now supports NumPy's *.npy format.
• Speedup of all routines that use the Fast Fourier transform (autocenter, align, ialign, itrack_peak, and kinematicsim) by 50%.

Release 2.1.1: autocenter

This release brings one major change:

• Added the :func:autocenter routine, to automatically find the center of diffraction patterns. This works for both single-crystal and polycrystalline patterns.
  as well as one deprecation:
• support for Python 3.6 and NumPy<1.17 has been dropped

Release 2.1.0: infrastructure improvements and various fixes

This release brings major infrastructure improvements, which in turn have allowed to squash some bugs.

• Migration of continuous integration testing to GitHub Actions.
• Migration of test infrastructure to pytest.
• Tests are now included in source distributions.
• Code snippets in documentation are now tested for correctness.

User-facing changes:

• Fixed an issue where the detected shift in skued.align and skued.ialign might be partial (i.e. only shift in one direction).
• Fixed an issue with packaging data on Linux.
• The fast argument to skued.align and skued.ialign has been removed. It was previously-marked as deprecated.
• Added pre-emptive support for Python 3.10+ by removing deprecations.
• Increased the precision of the pseudo-voigt approximation in skued.pseudo_voigt.
• Fixed many issues regarding documentation being out-of-date.

Release 2.0.6: Python 3.9 support and relicensing

This release sees the following changes:

• scikit-ued is being re-licensed from the MIT license to the GPLv3 license.
• The fast argument to skued.align and skued.ialign has been deprecated. Its value has no effect anymore.
• Official support for Python 3.9.
• Removed explicit requirement for the tifffile package.

Release 2.0.5: kinematic diffraction simulation

This release brings two major additions:

• Added skued.kinematicsim, a simple function to compute electron diffraction patterns from
  crystals structures in the kinematic approximation (i.e. thin samples).
• Added the skued.RingArcSelection area.

as well as documentation improvement and fixes.