v2.2.0

Changes from v2.1.0

New Features

• sim-trhepd-rheed can now treat many-beam condition data [#56]
• Limitation (constraint) of the parameter space [#56]
  • Parameters $x$ satisfying the inequity $Ax+b \ge 0$ are considered
    • $A$ is a matrix and $b$ is a vector
    • This inequity $(\ge 0)$ means that all the elements are non-negative
• runner.mapping.b in the input file now accepts a column vector [#52]

Fixed issues

• In the PAMC, when a process has no walkers the program stops [#51]
• In the exchange MC, the too large number of MC samples crashes the program in the post process [#54]
• When the mesh file (grid file) has only one parameter point the program stops [#57]

v2.1.0

Release note

Update from v2.0.0

Samples

• Add a sample tool for generating an xyz file from the optimized atomic structure obtained by SXRD (#44)
• Simple sample for function solver is added (#40)

Documents

• Simple tutorial for adding user-defined solver is added (#37)
• sample/leed moves to sample/leed/mapper/ (#41)
• Description of solver.post.Rfactor_type for sim_trhepd_rheed solver is updated (#47, #48)

For developers

• Automatic test for building documents (#38, #39)

Misc

• Delete tool/todft directory (#42)
  • NOTE: todft tool has already been moved to sim-trhepd-rheed
• Remove old directory (#43)

Full Changelog: v2.0.0...v2.1.0

v2.0.0

Release Note

Update from v1.0.1

General

• rearranged the structure of samples directories (#33)
• change TOML python package from toml to tomli (#26)
  • toml is deprecated: we continue a support of this in this version but will drop it in future

Algorithm

• 🆕 pamc (#30)
  • population annealing monte carlo method
• exchange
  • becomes able to specify temperatures by inverse temperature beta (#27)
  • becomes able to calculate multiple replicas in one MPI process (#22)
  • becomes able to search the discrete parameter space as mapper and bayes algorithms via MeshData.txt (#21)
    • for MC update, it needs a neighbor list; py2dmat_neighborlist can generate it.
  • a post-processing script, separateT.py, is absorbed into the main program as a post process (#22)
  • when a replica goes out of bounds, the solver won't be called (of course, update will be rejected) (#28)
• 🆕 montecarlo (for developers) (#19)
  • base class for Monte Carlo methods
  • exchange and pamc are based on it
  • Downgrade NumPy.Random to the old API (#18)

Solver

• 🆕 sxrd
  • surface x-ray diffraction (SXRD) experiments
  • wrapper for sxrdcalc/sxrdcalc
• 🆕 leed
  • Low Energy Electron Diffraction experiments
  • wrapper for SATLEED
• sim_trhepd_rheed
  • Changed the name of the working directory (#33)
  • Added an option to remove old working directories (#33)

Full Changelog: v1.0.1...v2.0.0

v2.0-beta

Release Note

Update from v1.0.1

General

• rearranged the structure of samples directories (#33)
• change TOML python package from toml to tomli (#26)
  • toml is deprecated: we continue a support of this in this version but will drop it in future

Algorithm

• 🆕 pamc (#30)
  • population annealing monte carlo method
• exchange
  • becomes able to specify temperatures by inverse temperature beta (#27)
  • becomes able to calculate multiple replicas in one MPI process (#22)
  • becomes able to search the discrete parameter space as mapper and bayes algorithms via MeshData.txt (#21)
    • for MC update, it needs a neighbor list; py2dmat_neighborlist can generate it.
  • a post-processing script, separateT.py, is absorbed into the main program as a post process (#22)
  • when a replica goes out of bounds, the solver won't be called (of course, update will be rejected) (#28)
• 🆕 montecarlo (for developers) (#19)
  • base class for Monte Carlo methods
  • exchange and pamc are based on it
  • Downgrade NumPy.Random to the old API (#18)

Solver

• 🆕 sxrd
  • surface x-ray diffraction (SXRD) experiments
  • wrapper for sxrdcalc/sxrdcalc
• sim_trhepd_rheed
  • Changed the name of the working directory (#33)
  • Added an option to remove old working directories (#33)

v1.0.1

Release Note

Update from v1.0.0

Documents

• In tutorial for minsearch, now plot rocking curves in directories named ExtraLog*, not Log*
• Added description of base.root_dir
• Fixed some typos

v1.0.0

Release note

Changes from v0.1.x

New features

• Generate Meshdata ( #1 )
• Add new benchmark function, himmelblau (#3)
• Logging of callings solver (#4)

Changes

• Rename surf solver to sim-trhepd-rheed (#2)
• Rearrange documentation orders (#6)
• Change bayes input sections (#8)

Documents

• Add English manual (#7)

For developers

• Document for the topic branch with the name of '*-autodoc' will be deployed to https://issp-center-dev.github.io/2DMAT/manual/*/ (#5)

v0.1.0

2DMAT -- Data-analysis software of quantum beam diffraction experiments for 2D material structure

2DMAT is a framework for applying a search algorithm to a direct problem solver to find the optimal solution. As the standard direct problem solver, the experimental data analysis software for two-dimensional material structure analysis is prepared. The direct problem solver gives the deviation between the experimental data and the calculated data obtained under the given parameters such as atomic positions as a loss function used in the inverse problem. The optimal parameters are estimated by minimizing the loss function using a search algorithm. For further use, the original direct problem solver or the search algorithm can be defined by users.
In the current version, for solving a direct problem, 2DMAT offers the wrapper of the solver for the total-reflection high-energy positron diffraction (TRHEPD) experiment. As algorithms, it offers the Nelder-Mead method, the grid search method, the Bayesian optimization method, and the replica exchange Monte Carlo method. In the future, we plan to add other direct problem solvers and search algorithms in 2DMAT.

Document

• English (in preparation)
• 日本語

py2dmat

py2dmat is a python framework library for solving inverse problems.
It also offers a driver script to solve the problem with predefined algorithms and solvers (py2dmat also means this script).

Prerequists python and python packages

• Required
  • python >= 3.6
  • numpy >= 1.17
  • toml
• Optional
  • scipy
    • for minsearch algorithm
  • mpi4py
    • for exchange algorithm
  • physbo >= 0.3
    • for bayes algorithm

Install

• From PyPI (Recommended)
  • python3 -m pip install py2dmat
    • If you install them locally, use --user option like python3 -m pip install --user
    • pipx may help you from the dependency hell :p
• From Source (For developers)
  1. update pip >= 19 by python3 -m pip install -U pip
  2. python3 -m pip install 2DMAT_ROOT_DIRECTORY to install py2dmat package and py2dmat command

Simple Usage

• py2dmat input.toml (use the installed script)
• python3 src/py2dmat_main.py input.toml (use the raw script)
• For details of the input file, see the document.

License

This package is distributed under GNU General Public License version 3 (GPL v3) or later.

Copyright

© 2020- The University of Tokyo. All rights reserved.
This software was developed with the support of "Project for advancement of software usability in materials science" of The Institute for Solid State Physics, The University of Tokyo.