CHANGES.txt

* PowderModel: add some hints on higher memory consumption with more threads
* correct use of OPENMP in c-code and performance optimization
* fix a memory leak in the c-code (not freed memory of function pointers)
* improved error checking and clean up in c-code
* convert to meson-python bases build system

v1.7.10, 2025-01-17

* add Python 3.13 wheel generation, otherwise equivalent to v1.7.9

v1.7.9, 2025-01-16

* avoid use of scipy.misc package. make xrayutilities work with scipy>=1.15 (#202)
* fix typo in GetStress method of Material/Crystal.
* improve wording in Experiment class docstrings (#192)
* fix typo in cache key of atomic structure factor (#189)
* Use numpy>=2 as build dependency only for python>=3.9 (#188 by @t20100)

v1.7.8, 2024-06-05

* Numpy-2.0 support for wheels on PyPI
* fix an Exception in the PowderDiffraction calculation occuring for some
  specific parameters (#179, thanks to @mendenmh).
* allow different Miller indices for the various layers in a dynamic diffraction calculation

v1.7.7, 2024-04-19

* enable creation of Python 3.12 packages
* enable correct parsing of Sardana MCA data in spec files (by @dschick)
* fix for using lmfit-1.3.0

v1.7.6, 2023-11-16

* fix parsing error due to missing header entry in Rigaku ras file
* allow naming of simulation materials (Powder, Layer)
* example script for fitting diffraction data of an epitaxial multilayer (by @VeryBitter)
* update AlAs lattice parameters from APL 66, 682 (1995) (by @VeryBitter)
* enable building Python 3.11 wheels

v1.7.5, 2023-06-14

* Added AlSb to predefined_materials.py (#164) (by @andrews-tuwien)
* allow parsing of SPEC files/scans with missing data entry
* implement the calculation of the anisotropic Poission ratio of materials
  (#139). Related to that a method in the DarwinModel was renamed and declared
  internal.
* extend continuous integration tests by adding doctests
* directly depend on the lmfit package for implementation of our fitting
  models.
* more possibilities to specify filename lists in getras_scan, getSeifert_map,
  and getxrdml_scan
* allow bytes-stream instead of filename in XRDML
* enable setting free lattice parameters from the Crystal instance
* minor code improvements, forcing flake8 to pass upon merging a PR #149
* fix typo in Poisson ratio of GaInAs Darwin model (introduced in 1.7.4)
* make C-code respect the VERBOSITY. C-Code should now be silent with
  VERBOSITY<INFO_ALL. Otherwise lots of info is printed. #145
* remove set_bit/clear_bit functions in favor of using enum.IntFlag
* use plain pytest in the CI integration (no code change in the package)
* remove build_doc target for setup.py, use sphinx (see builddocs pipeline)

v1.7.4, 2022-10-31

* fix data type casting of cch1,2 in QConversion.init_area -> these can now be floating point values (#141 by @trdd)
* fix docstring for March-Dollase texture model in xu.simpack.Powder. The
  parameter preferred_orientation_factor was described wrong. Platy crystals
  need a factor < 1! see W. A. Dollase, J. Appl. Cryst. (1986). 19, 267-272
* fix Poisson ratio of GaInAs Darwin model (thanks to Danny Chrastina for
  reporting)
* Additional convenience functions for elasticity tensor symmetries for
  Monoclinic and Trigonal symmetry (#138 by @f-iniv)
* GetStrain and GetStress functions for xu.material.Material (#138 by @f-iniv)
* enable correct parsing of Seifert RSM map measurements
* matplotlib format bugfix in psd_chdeg
* improvements for kappa goniometers
* remove use of deprecated scipy namespace

v1.7.3, 2021-11-26

* enable building Python 3.10 wheels
* fix missing files in source package on PyPI
* make builddocs pipeline more robust

v1.7.2, 2021-11-22

* fix returned fitting error in area_detector_calib* (issue #126)
* remove dependency on soon deprecated distutils
* fix regression bug with matplotlib 3.5
* update copyright notices from git commit history
* fix code error found by linting (flake8)
* CIF atom type normalization of oxidation state (issue #122)
* improve documentation of PowerModel multiprocessing code (issue #121)
* fix numpy-1.20 deprecation warnings
* bugfix in kinematical diffraction model.
* add visualization of individual layer densities (issue #107)
* new coplanar_alphaf helper function

v1.7.1, 2021-01-02

* harmonize dependency versions between README and setup.py
* remove PDF documentation and other generatable content from the GIT
  repository
* new azure pipeline for documentation generation
* new azure pipeline for wheel generation (issue #111)

v1.7.0, 2020-12-08

* allow string decoding to fail in tif headers
* ignore nan values in maplog
* add Azure pipeline for continuous testing and automatic builds
* performance optimizations in Crystal.environment
* Continous integration tests running on MacOS, Windows and Linux
* new flavor of material/Crystal definition from fractional atomic coordinates
* bugfix in derivative functions (xu.math.functions)
* effectiveDensity slicing for XRR simulations, thanks to MM
* new possibility to transform unit cells and find symmetries
* bugfix of numpy.set_printoptions use
* require Python 3.6 due to use of f-strings
* new PowderModel.plot method to quickly visualize a Powder pattern
* remove deprecated fit_xrr function
* SGLattice.get_allowed_hkl function
* make SGLattice.convert_to_P1 a normal method
* SGLattice improvements: add symmetry operations, equivalent_hkls,
  reflection_conditions, and allowed_hkl functionality
* fix SGLattice.isequivalent function to work for non-centrosymmetric
  materials
* implementation of the March-Dollase model for preferred orientation in
  PowderDiffraction
* Mayavi based visualization of unit cells: Crystal.show_unitcell
  -> new optional dependency on Mayavi
* fix new Pilatus TIFF files which seem to have \xff padding
* implement Gaussian beam shape for XRR modelling
* update documentation to recommend the use of pip for installation
* remove multGaussFit/Plot and set multPeakFit/Plot deprecated
  -> use the much more powerful lmfit-package instead!
* rename dummy element to "Dummy" and set sensible default parameters

v1.6.0, 2020-01-08

* use of tox to run unittests consistently in a virtual environment
* build element database during installation from sources
* new source layout which offers many advantages
  (https://blog.ionelmc.ro/2014/05/25/python-packaging/#the-structure)
* remove Python2 code path, improve code readability
* fix use of the agg matplotlib backend with FitModel
* from this point on Github master will be broken on Python 2

v1.5.4, 2020-01-07 final python2 compatible release

* fix issue #84: edge case problemes in XRDML parser
* matplotlib compatibility improvements

v1.5.3, 2019-10-09

* issue #87: fix matplotlib inline backend in FitModel.fit
* enable parsing of Panalytical XRDML version >2.0
* issue 79: fix parsing of Sardana generated spec files; thanks to DS
* issue 81: fix parsing materials with space group 47
* packaging cleanup: remove gitignore and other dev-files
* issue 77: complete multiprocessing cleanup upon PowderModel.close()
* issue 76: cif parser: fix bug introduced in version 1.5.2

v1.5.2, 2019-06-28

* fix bug in PowderDiffraction where the last Bragg peak before tt_cutoff was
  omitted, thanks to NG for reporting the issue
* new model for calculating signal of mosaic small crystals
* add arbitrary line cut function
* fix parsing of empty scans in spec files
* remove geth5_map function -> use geth5_scan instead
* support inf/nan values in spec files
* setup.py fixes for Debian packaging
* CIF export: make _atom_site_label unique
* change of SMaterial.density to a unit of kg/m3
* deprecate the fit_xrr function, which will be remove in the next major
  version -> use the new FitModel
* new FitModel class for fitting of instances of LayerModel
* fix bug in PowderDiffraction used with enable_simulation=False

v1.5.1, 2019-01-07

* simple unit cell visualization function: Crystal.show_unitcell()
* add new 'radius' and 'color' properties to Elements
* improve testing of various unit cell settings in CIFFile
* fix bug in handling negative data in kill_spike, thanks to reihel for
  reporting
* Panalytical XRDML: parse hkl from file if present
* CIF exporter and possibility to parse CIF files from string representations
* new reflectivity code based on transfer matrices (contributed by Numan
  Laanait, nlaanait@gmail.com)
* unittest and code quality updates
* show_reciprocal_space_plane function to plot a cut in reciprocal space
  together with the peak positionn of a given material
* new FourC experiment class, equivalent to HXRD, but using a four circle
  goniometer for consistent Q2Ang and Ang2Q functions
* add spline coefficients to PowderModel fits
* allow single Layer object as argument in LayerModel
* rewrite of line cut functions, shorter code, faster execution, more
  versatile
* avoid loops in Experiment.Q2Ang functions
* clean README and add new CONTRIBUTIONS instructions

v1.5.0, 2018-07-04

* improved docstring formating
* simulation code for diffuse X-ray reflectivity
* more flexibility in defining materials Wyckoff positions
* convenience function to calculate diffracted intensities of thin films to
  compare various Bragg peaks (coplanar diffraction geoemetry)
* simplify WyckoffBase of Alloy, by avoiding double entries
* CIF parser now is able to fully parse files without symmetry operations
* CIF parser now distinguishes multiple "_data" entries
* improvement of the scanning XRD dataset parsers for ID01 (FastScan)
* PEP8 improvements
* speed optimization in EDF parser
* add getheader_element to SPECScan
* try to parse oxidation state of elements from a CIF file, and ignore '.' and
  '?' values for Debye-Waller factors and occupation
* numpy 1.14 related fixes (mostly in file parsers)
* fix bugs in spacegroups whos Wyckoff position contains a multiplication
* make U/UB matrix compatible with the definition in SPEC
* remove deprecated Lattice/LatticeBase code
* update XRR example

v1.4.2, 2017-11-28 "The Praha edition"

* introduce convenience function HKL to convert q-vectors to HKL of materials
* update outdated information and fix typos in README
* introduce abstract classes where useful
* code cleanup (import statements)
* vastly improved CIF file support with recognition of correct space group and
  Wyckoff position
* use AppVeyor.com for Windows testing and binary creation
* make wavelength handling in PowderDiffraction more flexible
* fix bug in SiGe alloy class
* python2 fix for EDFFile which could have failed as described in issue #41
* fix clear_bit utility function
* fix float/int conversion in line cut functions for newer numpy

v1.4.1, 2017-04-03

* new materials (Cr, tetragonal BaTiO3)
* fix bug in DarwinModel
* fix multiprocessing on MS Windows platforms
* fix BUG in tetragonal crystals (wrong symmetry was used!)

v1.4.0, 2017-02-09

* fix of negative HKL directions in RelaxationTriangle
* new SGLattice class which replaces Lattice and allows for easier material
  definitions
* old Lattice class is still included but now deprecated and will be removed
  in >1.5.0
* new PowderExperiment, PowderDiffraction and PowderModel classes for proper
  simulation of powder patterns. Thanks to Marcus Mendenhall for his
  contribution of a Fundamentals parameter approach powder profile
  calculation.
* cache database results -> huge speedups in repeated calls with same
  parameters
* new file parser for ILL numor files (station D23, maybe more general)
* add possibility to fix the pixel size when calibrating 2D detectors
* new normalized Lorentzian function (NormLorentz1d)
* add fully asymmetric Pseudovoigt (different width and eta)
* remove unphysical variation of structure factor in kinematical 2-beam model
* new dynamical multibeam simulations based on the Darwin theory of dynamical
  diffraction
* include pip install guide into readme
* fix default wavelength selection in line cut functions.

v1.3.3, 2016-08-24

* fix packaging of the README.md file

v1.3.2, 2016-08-24

* call PowderIntensity from xu.Powder init routine
* fix problems in Crystal class when no lattice base is defined
* add parsing for simply chemical formulas in Amorphous definitions
* fix setup.py issue related to test installation
* fix bug in fuzzy gridders (first data point was not always treated correct!)
* added absorption_length method in all material classes

v1.3.1, 2016-05-19

* fix windows build due to lack of support of C99 in MSVC

v1.3.0, 2016-05-19

* change to setuptools for installation/package building
* major code cleanup (remove of duplicated code using Clone Digger)
  http://clonedigger.sourceforge.net
* bug fixes in Alloy initialization
* documentation building now requires numpydoc
* central version number in VERSION file (also available in the package after
  importing: import xrayutilities as xu; xu.__version__
* new examples using the new XRR and XRD simulation-code
* improve handling of SiGe and AlGaAs alloy: alloying on atomic level
* add relaxation parameter to PseudomoprhicMaterial to model partially relaxed
  material
* fix fwhm_exp for very narrow peaks
* fitting of x-ray reflectivity using the lmfit package.
  If you want to use this new feature you have to install lmfit
  see https://pypi.python.org/pypi/lmfit/
* new subpackage simpack for diffraction and reflectivity simulations from
  thin film samples. So far implemented are
  - a basic kinematical diffraction model
  - two dynamical diffraction models (simplified and general 2-beam theory)
  - a model for specular reflectivity calculations using a Parrat-like
    formalism
* fwhm_exp function was moved to math's subpackage! This was necessary in
  order to avoid circular dependencies, but anyhow the function also
  fits better there!
* improvements of peak_fit function
  - better ability to handle linear background
  - more stable fitting due to use of deriavitives (PseudoVoigt)
* minor improvement in peak_fit's plotting

v1.2.1, 2016-01-08

* Uiso parsing from CIF files for better structure factors
* support for parsing Rigaku RAS files (Smartlab Guidance)
* new getDetectorPos/Distance routines in QConversion
  This can be used for all types of detectors to determine their position and
  distance for data normalization purposes
* add FuzzyGridder3D class
* add FuzzyGridder2D class (see mailing-list for details)
* add possibility of Pseudo-Angle (basically arbitrary contraints) to Q2AngFit
* add Carbon(diamond), HOPG materials
* enhancement in FIO file parser
* much improved TIFF file parser
* IntensityNormalizer gained possibility to normalize CCD data

v1.2.0, 2015-08-03

* Radicon file support was removed (no test scripts available!)
* replace pytables by faster and more convenient h5py
* bug fixes/improvements in the file parsers (discovered by the unit tests!)
* new unittests for file parsers (spec, edf, xrdml, esg, nja, fio, cbf, tif)
* added different oxidation states of the chemical elements to the database
* huge speed improvements for the database access, especially when querying
  multiple values at the same time
* add PseudoVoigtAsym option in xu.math.peakfit
* memory usage optimization in FastScanCCD

v1.1.2, 2015-04-22

* new webpage theme to allow for easier reading (esp. mobile devices)
* huge set of improvements for working with FastScans (ESRF ID01)
* keep_open option in EDFFile to speed up reading of multiple images
* new convenience function for parsing data from spec-files
* improve peak position finding in area-detector calibration
* added new material AlAs

v1.1.1, 2015-01-22

* new PseudoVoigt1dasym function implementing an asymmetric PseudoVoigt peak
* FuzzyGridder1D: improved 1D gridder using a range of value for each data
  point
* fix memory leak in XRDMLFile by abandoning minidom
* edf-parser improvements: support for multiple images per file
* spec-parser improvements, keep file closed when not parsing it
* peak_fit: linear background feature
* peak_fit function: enhancement of automatic initial parameter determination
* improvements in the Material.environment function (occupancy awareness)
* reciprocal space conversion for goniometers with translation-axis in the
  detector arms (point, linear, area detectors)
* reciprocal space conversion of energy scans for point, linear, area
  detectors
* addition of serval test routines to improve the code quality

v1.1.0, 2014-10-09

* fix bug in Q2Ang with refrac = True (omega angle could have been reported
  incorrect and added some geometry checks/warnings)
* more flexibility in sample position/orientation definition
* documentation updates
* small bug fixes in matplotlib import code
* c-header-code cleanup
* reformat c-code to conform to PEP7
* reformat python code to conform to PEP8

v1.1.0-beta, 2014-09-16

* new 1D Gridder working consistent with 2D and 3D
* add easy Powder Diffraction cif file parser
* addition of several test routines to improve the code quality
* new mathematical functions in math subpackage
* fixes for possible segfault in the Gridder classes
* considerable reduce size of elements.db file
* FastScan parser for spec-mesh scans and FastScan/kmap data from ESRF 
  (thanks to RG for help with testing and implementation ideas)
* code-cleanup
* minimal bug-fixes in io (xrdml)
* documentation updates
* add environment function to Materials: function calculates the atomic
  neighbors around a given position in the unit cell
* documentation update regarding the Q2Ang and Q2AngFit functions
* fixed inconsistency bug in NonCOP and HXRD's Q2Ang function (thanks to FZ)
* SPEC-file parser improvements 
* python3.X compatibilty in single code-base

v1.0.6, 2014-04-15

* new way of accessing spec-scans when parsing the spec file
  new: spec scans can be directly accessed with there spec-scan-number
* sequential gridding: usage improvements
* sample displacement error consideration in qconversion (experimental)
* new Q2AngFit function able to fit the goniometer angles or a q-vector
* minor fixes in the CIF file parser
* add workaround for internal coordinate system definition in the experiment 
  class
* new getxrdml_scan function to read (multiple) XRDML-files 
* fixes for the getxrdml_map function
* add packaging instructions for WIN64

v1.0.5, 2014-02-09

* Documentation updates
* minor compiler warning fixes and small bug fixes (line cut routines, 
  EDF file reading)
* more flexible shape of data arrays for Ang2Q function
* minor improvements of the spec-file parser
* added KeepData option to Gridders in 1,2,3 dimensions (sequential gridding)
* implement possibility to read gzipped, bzipped data files 
  (available for most data file readers)

v1.0.4, 2013-12-22 "the final Linz edition"

* bugfix in GID experimental class: qconv argument so far not used correctly!
* enhanced fitting routines
* spec files can now be parsed while gzipped
* CBF file parser for detector images 
* compile fixes for Microsoft compilers
* compile fixes on Mac
* new predefined materials (NaCl, FeO, Fe3O4, Co3O4,...)
* some minor bug fixes

v1.0.3, 2013-10-25

* build_doc target for setup.py to help build docs (thanks to F. Picca)
* fix python setup.py --help
* add PseudoVoigt function in 1D
* add Lorentzian peak shape in addition to Gaussian in new function
  multPeakFit
* some minor bug fixes

v1.0.2, 2013-09-17

* add documentation source to tarball
* generate info file from documentation
* make c extension work with python3, and in general make XU more python3
  friendly
* fix bug in SiGe materials lattice_const_AB function

v1.0.1, 2013-09-14

* new way of parsing motor positions and counter values from EDF files
* some fixes and updates in the ID01 example
* added new predefined materials
* new Material method "distances" to calculate atomic distances in the unit
  cell
* some minor bug-fixes 

v1.0, 2013-07-22

* add new example with hotpixelkill variant
* fix unicode string errors on Windows

v0.99.2, 2013-07-18

* fix issue in python-c-extension on windows
* add new example
* add functions to fit multiple Gaussian to a spectrum
* add tests to released files

v0.99.1, 2013-07-03

* documentation updates
* function analysis.area_detector_calib2 renamed to area_detector_calib_hkl
* added function to remove spikes from 1d data arrays 

v0.99, 2013-07-01 -- planned to be the initial release on PyPI