https://github.com/ttadano/alamode - v.1.5.0

New

• Structural optimization at finite temperatures based on SCPH and QHA (see RELAX_STR tag, &relax, &qha, &displace, and &strain field for more details.)
• New way to distribute anharmonic force constants to periodic images (PERIODIC_IMAGE_CONV=1) is set to default. If you want to move back to the older convention, please set PERIODIC_IMAGE_CONV=0.

Fix

• Fix a bug in IALGO=1 for the SCPH mode

Improvement

• NONANALYTIC=3 (Ewald summation) is now compatible with MODE = SCPH
• Computation of SCPH is accelerated
• More efficient and robust VASP XML parser (requires lxml)

What's Changed

• supply arguments which are used as global variables in functions in t… by @nim-hrkn in https://github.com/ttadano/alamode/pull/87
• structural optimization based on SCPH and QHA. by @r-masuki in https://github.com/ttadano/alamode/pull/136
• Faster mode transformation in calculating reciprocal-space IFCs. by @r-masuki in https://github.com/ttadano/alamode/pull/169
• Optimized PERIODICIMAGECONV=1 by @r-masuki in https://github.com/ttadano/alamode/pull/170
• Add tutorials by @r-masuki in https://github.com/ttadano/alamode/pull/171

Full Changelog: https://github.com/ttadano/alamode/compare/v.1.4.2...v.1.5.0

- C++
Published by ttadano over 2 years ago

https://github.com/ttadano/alamode - ALAMODE version 1.4.2 (2022-09-28)

Fixes

• Fix a bug in group velocity when NONANALYTIC=3

- C++
Published by ttadano over 3 years ago

https://github.com/ttadano/alamode - ALAMODE version 1.4.1 (2022-06-22)

Fixes

• Fix a bug in xml writer in the previous version

- C++
Published by ttadano almost 4 years ago

https://github.com/ttadano/alamode - ALAMODE version 1.4.0 (2022-06-21)

New

• FC_ZERO_THR tag (alm) for solving #71
• STOP_CRITERION tag (alm) for stopping the cross-validation calculation when appropriate
• FC3_SHENGBTE tag (alm) for saving third-order force constants for ShengBTE format
• ZMODE tag (anphon) for computing mode effective charges
• UCORR and SHIFT_UCORR tags (anphon) for computing displacement-displacement correlation function
• extract.py can extract data from vaspout.h5 for VASP6

Changes

• Update the predefined values of atomic weights and isotope factors
• Add new lines before printing warning/error messages. #61
• Make non-analytic correction of dynamical matrix to be periodic in G
• The file suffix for phonon eigenvectors is .evec, .band.evec, or .mesh.evec depending on the input KPMODE.
• scph_to_qefc.py now accepts ibrav/=0 and .fc files containing the macroscopic field

Fixes

• Fix minor typos in the documentation pages

- C++
Published by ttadano almost 4 years ago

https://github.com/ttadano/alamode - ALAMODE version 1.3.0 (2021-10-01)

Changes

• For OpenMX, displace.py now generates atomic positions in Cartesian coordinate instead of the fractional coordinate.
• Refactor the codes for future major release (This release does not change the user interface.)

Fixes

• Fix a few bugs in the SCP class. The stability of the phonon frequency interpolation has been improved.
• Fix other minor bugs and typos in documentation and header parts of output files.

- C++
Published by ttadano over 4 years ago

https://github.com/ttadano/alamode - ALAMODE version 1.2.0 (2021-04-02)

New features

• CMake installation option
• FC_BASIS tag (alm) for better stability of force constant symmetrization
• NONCOLLINEAR tag (alm) for phonon calculations with noncollinear magnetism
• NMAXSAVE tag (alm) that controls the maximum order of anharmonic terms to be saved in a file
• LMODEL = adaptive-lasso (alm) that performs adaptive LASSO
• KAPPA_COHERENT tag (anphon) for computing the coherent part of thermal conductivity
• ANIME_FRAMES tag (anphon) that controls the number of frames saved for animation outputs
• extract.py can extract the dielectric tensor and Born effective charges from vasprun.xml (VASP) and ph.x outputs (Quantum ESPRESSO) by the --get born option.
• fc_virtual.cpp which perform a virtual crystal approximation (VCA)-like interpolation of force constants

Changes

• The old tags DFILE and FFILE are now obsolete. Use DFSET instead.
• The filename extension of PREFIX.enet_cv is changed as PREFIX.cvset.
• CV_MINALPHA and CV_MAXALPHA are now set automatically (by default)
• The default value of CV_NALPHA is changed to 50
• The header part in PREFIX.evec has been modified slightly. Please be careful if you are using PREFIX.evec for further analyses.
• plotband.py now works nicely for discontinuous BZ paths.
• When BCONNECT > 0 and KPMODE = 1, phonon velocities, polarization vectors, and Grüneisen parameters are also reordered before saved in files.
• Support command line usage of dfc2

Fixes

• The tetrahedron method (ISMEAR = -1) now works correctly even when the number of momentum points along each axis is only one. (fix issue #10)
• Fix an issue of MPI communicators when sending large messages (> 2^31-1).
• The parser for LAMMPS now can read *.lammps file that contains charge entries. (fix issue #13)
• Fix a bug in the ANIME option

- C++
Published by ttadano about 5 years ago

https://github.com/ttadano/alamode - ALAMODE version 1.1.0 (2019-05-01)

New

• An interface to OpenMX code (contributed by Yuto Tanaka)
• Compressive sensing approach (LMODEL = enet) in alm code. Many new variables related to compressive sensing are also added. See the documentation page for details.
• SPARSE and SPARSESOLVER tags in alm code
• DOS-tag in anphon code
• A python script scphtoqe.py that converts the result of an SCPH calculation to Quantum ESPRESSO force-constant format.

Changes

• Default value of ICONST is changed to ICONST = 11
• Python scripts now work with python3 as well as python2
• Python interface scripts are moved to tools/interface
• Default values for MASS- and ISOFACT-tags are implemented
• Implement a sparse version of rref, which improves the performance of alm significantly.
• Performance improvements of anphon code.
• DFILE and FFILE in alm code are now deprecated. Use DFSET instead.
• &fitting field in alm is replaced with &optimize field.

Fix

• Fix a minor bug in calcdampingtetrahedron. The phonon linewidths at high temperatures and the thermal conductivities were not affected by this minor error. In a very low-temperature region (< 10 K), the thermal conductivity may have been underestimated.
• Fix other minor bugs

- C++
Published by ttadano about 7 years ago

https://github.com/ttadano/alamode - ALAMODE version 1.0.2 (2018-1-29)

New

• Phonon band connection by eigenvector similarity (BCONNECT tag)
• New option to turn on/off the symmetrization of Born effective charge (BORNSYM tag).

Changes

• Improve the performance of the "suggest" mode for hexagonal systems
• Use <unorderd_set> instead of <set> for better performance

Fix

• Fix a bug in the symmetrization of the Born effective charge

- C++
Published by ttadano over 8 years ago

https://github.com/ttadano/alamode - ALAMODE version 1.0.1 (2017-11-21)

Fix

• Fixed a minor issue in the previous version

- C++
Published by ttadano over 8 years ago

https://github.com/ttadano/alamode - ALAMODE version 1.0.0 (2017-11-21)

New

• Self-consistent phonon calculation (MODE = SCPH)
• Ewald summation for the non-analytic term of the dynamical matrix (NONANALYTIC = 3).
• Support of the CLASSICAL option.
• Python auxiliary script for LAMMPS

Changes

• P+ and P- are printed seperately in PREFIX.sps when SPS = 1
• Use C++11 standard. From this version, the C++ compiler must support the C++11 standard.
• The anphon code symmetrize the Born effective charges

Fix

• Loosen the tolerance to detect the multiplicity of force constants. This is an important fix for low symmetry structures.
• Fixed a problem of the restart mode of MODE = RTA

- C++
Published by ttadano over 8 years ago

https://github.com/ttadano/alamode - ALAMODE version 0.9.8 (2016-7-14)

This release contains some important bug fixes and performance improvements. Upgrading to the new release is recommended to all users of previous versions.

New

• New tag HESSIAN in program alm for printing entire Hessian matrix
• New tag KAPPA_SPEC in anphon for calculating spectra of thermal conductivity
• New option --offset in extract.py for subtracting residual forces (displacements) in an equilibrium structure from training data sets. We recommended to use this option if internal coordinates of atoms (Wyckoff potisions) have free parameters.
• New option --isotope in analyze_phonons.py

Changes

• Improve the performance of thermal conductivity calculations with the tetrahedron method (ISMEAR=-1). The new version is more than 3 times faster than the previous version.
• Improve the efficiency of the algorithm for generating constraings for the translational invariance
• Avoid 'NaN' in thermal conductivity calculations with imaginary branches
• Loosen the default value of TOLERANCE for detecting crystal symmetry
• Stop printing the CLASSICAL entry in PREFIX.result files
• Change the unit of the smearing width written in PREFIX.result files (a.u. --> cm^-1)

Fix

• Fixed an issue regarding the phonon-isotope scattering rate
• Fixed a bug in relaxation.cpp regarding the permutation symmetry of q mesh
• Fixed an invalid memory reference in analyze_phonons.cpp
• Added a routine to check the consistency of the crystal structure when the FC2XML tag is used in anphon

- C++
Published by ttadano almost 10 years ago

https://github.com/ttadano/alamode - ALAMODE version 0.9.7 (2016-3-10)

New

• New option SPS = 2 for anphon
• New tag MAGMOM for considering collinear spin (alm). The format is same as VASP.
• ALAMODE logo

Changes

• From this version, the variable 'NNP' for the translational part of space group operations will no longer be used. Because of this, file formats of SYMMINFO and SYMMINFO_PRIM also changed.
• Comment lines in the &cell, &position, and &kpoints fields are supported.
• Any words starting from 'N' or 'n' can be used instead of 'None' for the &cutoff field.

Fix

• Fixed a bug in fitting.cpp (alm) regarding out-of-range of an array.
• Fixed a bug in the calculation of the maximum distance for multi-body interaction clusters.
• Fixed a bug in the calculation of constraints for translational invariance. This bug appeared in previous versions with some specific structures/cutoff radii.
• anphon code incorrectly printed RMSD instead of MSD in previous versions. Fixed.
• qe2alm.cpp now works with ibrav=0.
• Fixed several issues in displace.py and extract.py for processing QE files.
• PRINTPR now works properly when KPMODE = 0.

- C++
Published by ttadano over 10 years ago