Abics

What's Changed

New features

* New script `abics_postproc` to calculate expectation values for each temperature (https://github.com/issp-center-dev/abICS/pull/67)
* New option for generating temperature points as beta space (https://github.com/issp-center-dev/abICS/pull/71)
* New machine learning energy solvers: Nequip and MLIP-3 (https://github.com/issp-center-dev/abICS/pull/68, https://github.com/issp-center-dev/abICS/pull/73)
* New option to set energy reference (https://github.com/issp-center-dev/abICS/pull/72)

Misc

* Drop Python 3.7/3.8 (https://github.com/issp-center-dev/abICS/pull/74)
* Available versions of pymatgen is also changed to ">=2022.1.20"

**Full Changelog**: https://github.com/issp-center-dev/abICS/compare/v2.1.0...v2.2.0

What's Changed

New feature

* Monte Carlo Sampling
* Population Annealing Monte Carlo (49 57)
* Grand canonical sampling (65)
* Users can customize how to calculate observables of configurations by defining Python function (64)
* Expectation values are calculated for each temperature in the RXMC sampler (50)
* Users can define arbitrary module and function for constraining (53)
* The initial configuration can be set (55)
* New input parameter `sampling.kTs` to specify temperature points directly (62)
* New solvers
* aenetPyLammps (56)
* invoke aenet solver as a library via LAMMPS (`pylammps`)
* Potts (46)
* Its primary purpose is to test algorithms such as sampling
* Logging (51 60)

For developers

* The format of the class names is unified (45)
* Old names are still available
* introduce factory for solvers (54)

Bugfixes
* Fix how to use `structure.to` (47 48)
* Fix `config.defect_structure.groups.coords` in `st2abics` (61)
* Fix `ignoresp` in AL (58)

**Full Changelog**: https://github.com/issp-center-dev/abICS/compare/v2.0.1...v2.1.0

<a name ="release_jp">Release note

[ 日本語 / [Eng](release_eng) ]

<a name ="release_jp">Release note

[ 日本語 / [Eng](release_eng) ]

- Created three new scripts: `abics_sampling` for sampling, `abics_mlref` to create training data for machine learning, and `abics_train` to train a neural network configuration energy model using aenet.
- Changed input file format to allow parameters to be set for each script.
- Updated manual and tutorial.

[Forum (GitHub issue)](https://github.com/issp-center-dev/abICS/issues)

<a name ="release_jp">Release note

[ 日本語 / [Eng](release_eng) ]
[abics-v1.0.0.tar.gz](https://github.com/issp-center-dev/abICS/releases/download/v1.0.0/abICS-v1.0.0.tar.gz)にはマニュアルのpdfファイルが入っています。"Source code"ではなくこちらのファイルをダウンロードしてください。