v0.5.1

This is mainly a maintenance release bringing up to date support for the project dependencies and revamping the testing infrastructure.

What's Changed

• [lammps] Fix handling types when restoring by @pabloferz in #344
• Bring CI up to date by @pabloferz in #348
• Improve support for jax.example_libraries.optimizers by @pabloferz in #350
• Improve compatibility with jax by @pabloferz in #351

Full Changelog: v0.5.0...v0.5.1

v0.5.0

What's Changed

Backends

• [LAMMPS] Add conversion factor depending on the unit system by @pabloferz in #325

Saving and Restarting Simulations

• Count how many times each method is called by @pabloferz in #292
• Add functions for saving and loading simulation results by @pabloferz in #304
• Fix incorrect ordering of the ase cell by @yzjin in #306

Sampling Methods

• Allow different number of grid points along each axis by @pabloferz in #300
• Sirens sampling method by @pabloferz in #330
• Add options for abf to use pseudo inverse from numpy by @yihengwuKP in #334
• Fix padding width by @pabloferz in #333

Collective Variables

• Implementation of eRMSD (AA), eRMSD (CG) and RMSD (Kabsch) [all unweighted] by @yihengwuKP in #321

Examples and Documentation

• Fix colab link and improve wording by @pabloferz in #288
• Add spectral abf example to lammps by @gustavor101 in #332
• Upgrade HOOMD-blue and OpenMM in Colab by @pabloferz in #341
• Update notebooks to work with latest version of the precompiled Colab environment by @pabloferz in #342

Dependencies Compatibility

We try to maintain backward compatibility with our main dependencies to make sure that users are able to run in as many computational environments as possible.

• Compat option for jax devices by @pabloferz in #320
• Fix plum compatibility issue caused by 2.3.0 release by @yzjin in #324
• [compat] Handle deprecation of jax.Array.device_buffer by @pabloferz in #336
• Compat option for jax._src.core.canonicalize_shape by @pabloferz in #340

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

v0.4.0

What's Changed

• Enable custom bias classes for umbrella integration (#177)
• Add a take_snapshot method to backend Samplers (#190)
• Allow multi component CVs (#192
• Flexibility in choosing which orientation the acyclindricity is calcualted (#194)
• Add support for HOOMD 3 (#200)
• Support jax-md as backend (#199)
• Add restraints in metad.py for grids (#217)
• Phase angle CV (#227)
• Add GeM CV (#242)
• Interface for restarts (#254)
• LAMMPS backend (#274)
• Revamp documentation (#281)

And many more bug fixes and improvements!

Full Changelog: v0.3.0...v0.4.0

v0.3.0

This new release includes usability enhancements, bugfixes and improvements over these previously existing methods:

• ABF
• ANN
• FUNN
• Harmonic biasing
• Umbrella integration
• Metadynamics
• Forward Flux Sampling

And these new sampling methods:

• CFF
• SpectralABF
• Well-tempered Metadynamics

ASE is now a supported backend which brings integration with many MD and QM/MM code bases including LAMMPS, CP2K, Quantum Espresso, VASP, among many others.

Other important key additions, include:

• A new interface to allow running enhanced sampling simulations on multiple GPUs. For instance, different windows in UmbrellaIntegration can be sampled in parallel on different GPUs, the same is the case for multiple replicas for most sampling methods in PySAGES.
• Many example scripts and Google Colab notebooks have been added as introductory resources to PySAGES.
• A user interface to easily analyze the results of a sampling run, enabling users to easily obtain Free Energy profiles.

v0.2.0

This new release includes usability enhancements, bugfixes and improvements over these previously existing methods:

• ABF
• ANN
• FUNN

And these new sampling methods:

• Harmonic biasing
• Umbrella integration
• Metadynamics
• Forward Flux Sampling

In addition, many example scripts and Google Colab notebooks have been added as introductory resources to PySAGES.

v0.1.0

First public release