Many thanks to Vahideh Alizadeh (@V-Alizade), Benedikt Bädorf (@benbaed), Marvin Friede (@marvinfriede), Thomas Froitzheim (@thfroitzheim), Igor Gerasimov (@foxtran) Hagen Neugebauer (@haneug), Alexander Paul (@apaulTQ), and Pit Steinbach (@pitsteinbach) for contributing to this release.

New Features

• Full ALPB/GBSA with empirical parameters by @haneug in #142
  • use --alpb for Analytical linearized Poisson-Boltzmann (ALPB) method, parametrized
  • use --gbsa for Generalized Born with Surface Area (GBSA) method, parametrized
  • use --gb for Generalized Born (GB) method, dielectric only
  • use --gbe for Generalized Born epsilon (GBε) method, dielectric only
  • use --cpcm for Conductor like polarizable continuum model (CPCM) method, dielectric only
• ASE support for solvation by @thfroitzheim in #241
  • Use TBLite(..., solvation=("alpb", "water")) for adding solvation models
• Density based features aka xtbml by @pitsteinbach in #224
  • use tblite --post-processing xtbml for computing xtbml features and saving them as npz file
• Add support for reading and writing npy and npz files by @awvwgk in #252, #256, #272, #277
  • use tblite --restart for writing the wavefunction as npz file
  • currently using restart is opt-in and will become default next release
  • results from singlepoint calculations in C and Python can be saved and loaded to npz
• Updated spGFN2 spin constants by @benbaed in #155, #274

Bug fixes

• Prevent calculator.get from printing properties by @apaulTQ in #221
• Fix creating efield interaction twice by @marvinfriede in #234

Documentation Updates

• Solvation documentation by @thfroitzheim in #226
• Add tutorials on using Python package by @awvwgk in #242
• Update singlepoint documentation by @V-Alizade in #251
• Add tutorial for the geometry optimization by @V-Alizade in #257
• Update documentation for ASE calculator by @awvwgk in #266, #267

Infrastructure & Refactoring

• Use C compiler for finding libraries by @awvwgk in #220
• Add TBLiteValueError to dict() method in Results container python by @pitsteinbach in #225
• Avoid division by zero by @foxtran in #228, #248
• Fill just allocated arrays with zeros by @foxtran in #230
• Remove unnecessary zeroing by @marvinfriede in #231
• Avoid possible problem with finalization in pure procedures by @foxtran in #232
• Bump setup-fpm to v7 by @marvinfriede in #236
• Fix debug compilation by @marvinfriede in #238
• Remove unused variables by @marvinfriede in #239
• Cleanup the xtbml imports by @thfroitzheim in #244
• Add Intel-LLVM support for meson by @foxtran in #255
• Fix CRC32 compilation with xtb: set proper sizes of integers instead default one by @foxtran in #254
• Fix CI by @foxtran in #258
• Use setup-fortran also for GCC by @awvwgk in #260
• Fix flicking overlap test by @foxtran in #259
• Exclude meson version 1.8.0 by @awvwgk in #265
• Deprecation warning for GCC 9 and adding compiler flag by @pitsteinbach in #264
• Update solver to directly return new density matrix by @awvwgk in #271

New Contributors

• @apaulTQ made their first contribution in #221

Full Changelog: v0.4.0...v0.5.0