Changelog

All notable changes to torch-pme are documented here, following the keep a changelog format. This project follows Semantic Versioning.

Unreleased

Added

• Allow passing full_neighbor_list and prefactor to tuning functions

Version 0.3.0 - 2025-02-21

Added

• Added support for Python 3.13

• Updated to metatensor-torch version 0.7

• Add a method to select a subset of a neighbor list based on a new cutoff (torchpme.tuning.tuner.TunerBase.filter_neighbors())

• Added an Ewald calculator for computing dipole-dipole interactions (torchpme.CalculatorDipole) using a dipolar potential (torchpme.PotentialDipole)

• Better documentation for for cell, charges and positions parameters

Removed

• Remove device and dtype from the init from all Calculator, Potential and Tuning classes

Version 0.2.0 - 2025-01-23

Added

• Added a PyTorch implementation of the exponential integral function

• Added dtype and device for Calculator classes

• Added an example on the tuning scheme and usage, and how to optimize the cutoff

Changed

• Removed utils module. utils.tuning and utils.prefactor are now in the root of the package; utils.splines is now in the lib module

• Tuning now uses a grid-search based scheme, instead of a gradient based scheme

• Tuning functions no longer takes the cutoff parameter, and thus does not support a built-in NL calculation.

• Refactor the torchpme.InversePowerLawPotential` class to restrict the exponent to integer values

Fixed

• Ensured consistency of dtype and device in the Potential and Calculator classses

• Fixed consistency of dtype and device in the torchpme.SplinePotential class

• Fix inconsistent cutoff in neighbor list example

• All calculators now check if the cell is zero if the potential is range-separated

Version 0.1.0 - 2024-12-05

Added

• First release outside of the lab