DOS interpolation

Resampling is now done by linear interpolation, which is more reasonable for continuous data and means there is not a minimal broadening requirement for a decent DOS output. The previous behaviour is more appropriate for irregularly-spaced data and remains available with the "spikes" option.

Bump to trigger Zenodo build

• Want to add a citation file...
• which needs some kind of permanent identifier like a DOI...
• but Zenodo needs a new version to be released before it will mint a DOI...
• but the next "thing to do" is add a citation file...
• so this release is identical to 0.4.0

New features

• GPAW calculations can now be imported if the gpaw python module is available, and (P)DOS plots generated.
• The new command-line tool galore-plot-cs generates plots of photoionization cross-section versus energy for a given set of elements

Major update including parametrised HAXPES data

Summary:

• BUGFIX: Yeh/Lindau weightings for partially-occupied orbitals
• BUGFIX: Odd behaviour in s orbitals including one transcription error
• Expanded weighting features: HAXPES data parametrised from Scofield tables
• Verbose output including cross sections, warnings and data source citations
• galore-get-sc tool for direct access to cross-section data
• Change Yeh labels from 'xps', 'ups', 'haxpes' to 'alka', 'he2', 'yeh_haxpes'

The "weighting" option can now accept floating-point energy values in keV. These will be used to determine cross-sections based on a fit to tabulated data (Scofield, 1973). The form of the fit is an 8th-order polynomial over the logarithms of energy and cross-section. Parameters are stored in an Sqlite database for rapid access.

The galore-get-sc tool allows easy access to this model when not examining a particular energy spectrum.

Core functionality, docs, testing in place

This code is now fairly usable. Projected density of states (PDOS) data can be read from generic files or VASP output, broadened, weighted by photoionization cross-sections and plotted with experimental data for comparison. Summary of changes in CHANGELOG, but the Tutorials section of documentation is a better tour of the features.

0.1.0

Initial alpha release, still not feature-complete

Added

• Initial prototype from random data
• Convolution algorithm used to apply Lorentzian broadening
• command-line interface
• Plotting to screen and file
• Limited internal tests
• setuptools-based distribution