Xrayutilities

* 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

* 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, nlaanaitgmail.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

* 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

* 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

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

* 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.