Changelog

All notable changes to this project will be documented in this file.

Version 3.0.0 - 2024-1?-??

New features

  • 4G-HDNNPs training/prediction in n2p2.

Version 2.3.0 - 2024-11-23

New features

  • Allow manual definition of train/test set (add set=train/test after begin in input.data).

  • Additional symmetry function log output, extra column spread.

  • Structure remap() function is now available in pynnp.

Maintenance

  • Fixed CI: Cython updated to version 3.0+, added __rmul__ to Vec3D in pynnp.

Important changes

  • Synchronized LAMMPS developer build with LAMMPS-distributed source: Finally switched to pair style hdnnp everywhere.

Documentation

  • Removed old nnp pair style description page.

  • Updated LAMMPS interface build instructions page.

  • Added train/test set split documentation.

Version 2.2.0 - 2022-05-23

New features

  • Normalization procedure on-the-fly via nnp-train, keyword normalize_data_set.

  • New normalization scheme based on force predictions.

  • New keyword force_energy_ratio for nnp-train.

  • Additional neighbor statistics output from nnp-scaling.

  • New tool nnp-checkdw (for testing purposes).

Important changes

  • New CI via Github Actions.

  • Use self-compiled doxygen for API docs, use version 1.9.4.

Bugfixes

  • Fixed missing atomic energy offsets for LAMMPS interface.

  • Fixed NaNs from nnp-dist when running mixed data sets.

  • Fixed bugs in pynnp examples.

Version 2.1.4 - 2021-05-11

New features

  • Changed build behavior for LAMMPS interface to conform with PR (lammps/lammps#2626)

Version 2.1.3 - 2021-04-26

Bugfixes

  • Fixed bug in nnp-scaling which could cause segfaults (MPI only).

Version 2.1.2 - 2021-04-12

New features

  • New tool to check numeric vs. analytic forces

Bugfixes

  • Fixed critical bug for large cutoff values (large compared to lattice vectors)

Version 2.1.1 - 2021-01-15

New features

  • Now compatible with latest LAMMPS stable release 29Oct2020

  • Improved EW warnings in LAMMPS

  • Citation header in screen output

  • Restructured training to prepare for multi-network HDNNPs (e.g. 4G-HDNNP)

  • New training “properties” can be added more easily now

Bugfixes

  • Fixed bugs in makefiles

  • Fixed bug when creating scaling data

Version 2.1.0 - 2020-12-21

New features

  • Polynomial symmetry functions (Martin P. Bircher and Andreas Singraber)

  • Better symmetry function cache mechanism

Important changes

  • Documentation updates (API docs in Sphinx via Breathe)

Version 2.0.3 - 2020-11-30

New features

  • Tool nnp-scaling puts out rejected configurations into a separate file.

Version 2.0.2 - 2020-11-25

Bugfixes

  • Bugfix in CabanaMD interface: SF group initialization

Version 2.0.1 - 2020-10-05

Bugfixes

  • Fix bug in makefile which sometimes causes a failed build due to parallel execution

Version 2.0.0 - 2020-10-05

New features

  • Training library (Multi-stream Kalman filter training).

  • Tools for HDNNP data set handling, etc.

  • Python interface (basic functionality).

  • Sphinx documentation (+ Doxygen API reference).

  • CabanaMD interface (by Saaketh Desai and Sam Reeve)

Important changes

  • License change from MPL 2.0 to GPL v3 or later.

Version 1.0.0 - 2018-08-13

New features

  • Core library (NN, symmetry functions, …).

  • LAMMPS interface.

  • Documentation (in parts) via doxygen.