Release Release 2.1.0 · Julian-Hochhaus/LG4X-V2

What's New (Major changes)

The fitting process was moved to its own thread, thereby enabling us to kill the fitting process from the GUI if the optimization takes too long. Thereby implementing the long-standing suggestion by @Hexanders in #5. Implemented in a552ec5
All user-defined models were moved to an external package called lmfitxps, which is available via pip and on github.
@kevinsmia1939 created a logo and added several files (desktop file, metainfo) in preparation for publishing the software on flathub in #79

Tables are now members of the class EditableHeaderTableWidget, which inherits from QtWidgets.QTableWidget and enables the user to edit table headers. Thereby, components used in the fit can now be renamed. In addition, clicking on the table header now opens a dialog, where the user can choose to remove selected components from the fit model. If header names are changed, the program checks for duplicates and renames components if necessary.
Implemented by @Julian-Hochhaus in 94f72a3, 3bad8ce, and ddd4364
Names of components are now exported to the pars.dat file so that the user is able to individually name fit components and later reuse the saved parameters belonging to specific identified and named fit components as suggested by @kevinsmia1939 and implemented by @Julian-Hochhaus in 90b466b.
Further improved the support for .vms files. Importing more information from VAMAS, such as excitation energy. See 5564c6b
The static Shirley background is now calculated using the same Model as the active one as suggested by @kevinsmia1939 . In addition, the static backgrounds are now exported as well and are displayed in the plot too. See Issue #66 and commits 369940b, 666b403, and 88d1c3f

Fix ValueError if - is used in input fields by @kevinsmia1939 in #50
Made sure, that the bg_mod variable is always set to None and caught all cases when bg_mod is not used in the model, solving Issue #52 reported by @kevinsmia1939. Fixed by @Julian-Hochhaus in 7532e1d.
zeros and NaN in the input data and in ModelResult are now caught, preventing the program from crashing as reported by @kevinsmia1939
in #55 and fixed by @Julian-Hochhaus in a77a20a
Pressing Refresh in the periodic table caused crashes due to a small typo. Issue #58 resolved by commit 21035237
removing constraints on fit parameters to enable the use of kinetic energy scale as well as binding energy scale, limits for the asymmetry parameter are now set based on which energy scale is used (see Issue #71, #59, and #73 by @kevinsmia1939 and commits 124e375, 884631f, and 164385a)
several bugs with the calculation of FWHM's as well as areas using binding energy scale were solved by @kevinsmia1939 and @Julian-Hochhaus , see 24d60f3, 9d021e8, and ea8ffb8

@kevinsmia1939 made his first contribution in #50. In addition, he put outstanding effort in improving the program by submitting various suggestions and issues as well as contributing to the code. In addition, he designed a logo and prepared the publication as flatpak. Thanks for your effort!

Full Changelog: 98c992e...70f37b1