Pdielec

A -threads or --threads option has been added to pdgui.
Added QT6 support by including the qtpy module
Changed the use of scipy.interpolate.trapz to np.trapz
Introduced a pyproject.toml file
Added version number to the pyproject.toml file and modified the setup.py file accordingly
Removed setup.cfg file. setup.py no longer sets the version number.
Updated the requirements.txt file using pipreq
Updated the conda_installs.txt file
Removed the need for the mkl package. But this means the number of processors used for threading
is set by the environment variables. It is best to NOT use threading and have the number of
processors set to the number of cores. The number of threads used by each core should be 1.
To force the BLAS routines to use threads it is best to use environment variables to control the
number of threads. For example....
export OMP_NUM_THREADS=1
export OPENBLAS_NUM_THREADS=1
export MKL_NUM_THREADS=1
export VECLIB_MAXIMUM_THREADS=1
export NUMEXPR_NUM_THREADS=1
The exporting of these variables is now performed automatically at runtime from App.py, before numpy is loaded
pdmake.py has been modified to allow -cpus and -threads to be passed to pdgui during benchmarking
Added additional escape characters to pdmake.py - to remove SyntaxWarning errors
Added additional escape characters to Mie.py - to remove SyntaxWarning errors
Added additional escape characters to CastepOutputReader.py - to remove SyntaxWarning errors
Added additional escape characters to CrystalOutputReader.py - to remove SyntaxWarning errors
Added a -style default in the creation of the App. Default style is "fusion"
Fixed a bug in the save cif file (ViewerTab)
Added xml readers to VaspOutputReader and QEOutputReader
Added a new example to QE/Urea which tests the xml reader
Added a new example to Vasp/Urea which test the xml reader
GenericOutputReader:
Created a new routine to find the frequencies and normal modes from the mass-weighted hessian
QEOutputReader:
Added a code to read pwscf.xml and tensors.xml
PhonopyOutputReader:
PDGui now reads the conversion factor store by Phonopy to scale the Hessian
Utilities:
Altered the logic in the "find_program_from_routine". The routine now
recognises pwscf.xml and vasprun.xml files. If an OUTCAR (Vasp) or a QE .log/.out is specified
on the command line the old method of analysing the output is used and the xml files are not read.
If vasprun.xml (Vasp) or a .dynG file is given on the command line then the xml will be read if it is there.
Changed the GTMCore.py use of clongdouble to cdouble.
Added a .nojekyll file into the docs/ directory to stop git hub pages over processing things
Changed the name PDielec to pdielec in setup.py to help with Windows install on conda-forge
Changed the find_program_from_name() function to return (program,qmprogram). Affected routines are;
- Utilities.py, HelperRoutines.py p1reader.py, preader.py, vibanalysis,py, App.py, NoteBook.py,
- MainTab.py, p2cif.py.
- Also changes in Helper/ and Jupyter/ Example directories.
Removed the slicing capability as it is no longer needed
Added new code for handling Phonopy DFT programs other than Vasp
The materials database is now cached to improve performance
The materials database is now kept open, instead of repeated opening and closing
Added support for AbInit v10 files, changed the AlAs abinit example to a version 10 example
Added support for Crystal 23: modified file extension handling to include .log as a crystal output file.
- Added a Crystal23 example in Examples/Crystal/Quartz
- Added an AbInit 10.0 example in Examples/Abinit/AlAs
- Added an QE 7.3.1 example in Examples/QE/Na2SO42-v7
- Added support for Phonopy-Crystal calculations: example in Examples/Phonopy/Crystal
Revamped the README.md files in the Examples directories
Modified the jupyter lab notebooks to use os.path.join for all directory access
Removed linked directories in Examples/Helper/Jupyter/..

The documentation has been revamped. An API description is now generated from each class and routine.
The examples of Jupyter notebooks has been added to.
Static routines have been created in the FitterTab to allow their use in notebooks
The UnitCell and SuperCell classes have been enhance to use a units specifier
Added new routines to the Materials and DielectricFunction classes to allow sigmas, oscillator strengths and frequencies to be changed
The write_cif() routine in UnitCell has been modified slightly to allow files to be written to

Modifications to the plotter tab to allow the use of wavelength and frequency units
Redesign of the single crystal scenario tab. The new design means that scripts are not backwards compatible.
- The mode setting has altered meaning that backward compatibility is broken.
The previous options are handled by the layer
mode can now be 'Transfer matrix' or 'Scattering matrix'
The default mode is 'Scattering matrix'
- Thick slab is now an option for an individual layer and has been termed 'Incoherent (non-reflective)'
- Other incoherent options have been added for a layer, using either intensities or phase matching
- the pygtm module has been modified so that subclasses are used for coherent and incoherent layers
and 'Transfer matrix' and 'Scattering matrix' classes are subclasses of the pygtm 'System' class
- removed the need to have specific super- and sub-strate entries in the singleCrystalTab
superstrate is the top of the material list and substrate is at the bottom.
- the old 'Thick slab' option is still available if the dielectric layer is the last in the list of materials
Updated the progress bar algorithms
Fixed a bug in the specification of the support matrix for the powder case
Fixed a bug in the reporting of absorptance, s- and p- were reported the wrong way round
Fixed a bug in the powder scenario tab, the super- and sub-matrix dielectric functions,
were not being set properly for frequency dependent permittivities
Fixed a bug in the plotting tab when writing to the spread-sheet a scalar permittivity
The option to use slicing has been removed
The option to do partial incoherence has been removed from the GUI but left in the code

Introduced various changes to handle the difference between Windows and Linux. The mainTab settings now has a compatibility mode option which can be 'Linux' or 'Windows'
Modified the AlN test case to use Windows compatibility mode
Added a slicing option in the single crystal case to address issues with exponential overflow
Added frequency dependent support matrix capability to the powder simulation
Added complex permittivity to the support matrix
Added incoherent light to the single crystal simulation
Added both partially incoherent thin film mode and incoherent thin film mode
Included smoothing for the partially incoherent case
Modified the handling of changing the legend in a scenario so a recalculation is not needed
Updated the documentation to reflect the new options for incoherent light
Added a range of units to handle crystal thickness
Changed the complex matrices in pyGTM to clongdouble
Modified pyGTM to try and improve its numerical stability
Added the concept of a slicing to the single crystal scenario, again attempting to improve numerical stability
Modified the reading of scripts as updates to python seem to have changed the exec() command, now read the whole file at once.
Add the concept of layers in the single crystal scenarios.
This allows multiple layers to be treated between the superstrate and substrate
Added a materials library which can be used to store material properties
The materials library can be used for powder and single crystal calculations
There are compatibility issues with previous version, so there may be some scripts which are not compatible.

Addressed problems after changes in numpy. Some np.int declarations were replaced with int

Addressed problems after changes in numpy
numpy.float and numpy.complex now replaced with float and complex