document

CrySPY is a crystal structure prediction tool written in Python.
CrySPY automates the following:

Structure generation
Submitting jobs for structure optimization
Collecting data for structure optimization
Selecting candidates using machine learning

License

CrySPY is distributed under the MIT License
Copyright (c) 2018 CrySPY Development Team

Latest version

CrySPY 0.10.0 (2021 July 25)

News

Click to expand old news.

[2021 July 25] version 0.10.0 released
- Support PyXtal 0.2.9 or later
- LAQA can be used with QE
- Upper and lower limits of energy for EA and BO
[2021 July 13] paper published
- Our paper on CrySPY software has been published in STAM:Methods
[2021 March 18] version 0.9.2 released
- Support pymatgen v2022.
[2021 February 7] version 0.9.0 released
- Interfaced with OpenMX
- Employ PyXtal library to generate initial structures
- If you use PyXtal (default), find_wy program is not required
- LAQA can be used with soiap
- Change the name: [lattice] section –> [structure] section
- Several input variables move to [structure] section
  - natot: [basic] –> [structure]
  - atype: [basic] –> [structure]
  - nat: [basic] –> [structure]
  - maxcnt: [option] –> [structure]
  - symprec: [option] –> [structure]
  - spgnum: [option] –> [structure]
- New features
  - Molecular crystal structure generation
  - Scale volume
[2020 March 19] paper published
- Our paper on adjusting the descriptor for CSP Bayesian optimization has been published in Physical Review Materials
[2020 February 16] version 0.8.0 released
- Migrate to Python 3
- CrySPY logo created
- Change several variable names and data formats
- Change style of output for energy: eV/cell –> eV/atom
- IDs of working directories corresponds to structure IDs
- New features
  - recalculation
  - manual select in BO
[2018 December 5] version 0.7.0 released
- New features
  - Evolutionary algorithm
[2018 August 20] version 0.6.4 released
[2018 July 2] version 0.6.3 released
[2018 June 26] Version 0.6.2 released
[2018 March 1] Version 0.6.1 released
[2018 January 9] paper published
- Our paper on CrySPY has been published in Physical Review Materials

Code contributors

Tomoki Yamashita (Nagaoka University of Technology)
Nobuya Sato (Tokyo Institute of Technology)
Hiori Kino (National Institute for Materials Science)
Kei Terayama (Yokohama City University)
Hikaru Sawahata (Kanazawa University)
Shinichi Kanehira (Osaka University)
Takumi Sato (Nagaoka University of Technology)

Reference

CrySPY（software）
- T. Yamashita, S. Kanehira, N. Sato, H. Kino, H. Sawahata, T. Sato, F. Utsuno, K. Tsuda, T. Miyake, and T. Oguchi,
  “CrySPY: a crystal structure prediction tool accelerated by machine learning”,
  Sci. Technol. Adv. Mater.:Methods 1, 87 (2021). Link
Bayesian optimization
- T. Yamashita, N. Sato, H. Kino, T. Miyake, K. Tsuda, and T. Oguchi,
  “Crystal structure prediction accelerated by Bayesian optimization”,
  Phys. Rev. Mater. 2, 013803 (2018). Link
- N. Sato, T. Yamashita, T. Oguchi, K. Hukushima, and T. Miyake,
  “Adjusting the descriptor for a crystal structure search using Bayesian optimization”,
  Phys. Rev. Mater. 4, 033801 (2020). Link
LAQA
- K.Terayama, T. Yamashita, T. Oguchi, and K. Tsuda,
  Fine-grained optimization method for crystal structure prediction",
  npj Comput. Mater. 4, 32 (2018).. Link

Link

GitHub repo GitHub release CrySPY utility Google group