What’s new!

Latest version (after 2024-09-05)

Updated Boost to 1.86.0 (patched for NumPy>=2.0).
Updated NumPy to 2.0.
Moved ReGraph and Lattice classes and make_regr() function from mc package to main package (to make phase enumeration working without Monte-Carlo package)
Added ads_site.py utility for enumeration of high-symmetry adsorption sites.
Added adlenum utility for enumeration of adlayer structures (similar to Hart-Forcade enumeration [10.1103/PhysRevB.77.224115], but in 2D and considering symmetries of adsorption complexes; see also [10.1021/acs.jpca.0c00846])
Added enumpaths.py utility for enumeration of diffusion trajectories.
Added test coverage meassurement, and reshuffled the test kit.
Speed-up transposition of TensorLinearOperator (should speed up TRG solver solve_TRG()).
Added support of Transfer-Matrix method for the N=1 case.
Added support of 1-D and non-periodic lattices by make_metropolis() function
Improved numerical stability of the Transfer-Matrix solver (solve_TM()).
Improved CUDA version of TRG solver (autodetect memory limits and block sizes).
Numerous fixed, memory leaks, refactorings.

Added routines for enumeration of cells and adlayer structures
Added support of non-diagonal supercell in `joined_cells_lattice_task()` and in Monte Carlo simulations
Added `sym_prec` key in AC samples for individual precission symmetry control.
Added Kinetic Monte Carlo support
Fixed incompatibility with spglib v.1.5. Please install old version spglib==1.14.1.post0 for pre 1.0 versions of SuSMoST
Upgraded supported Python version to 3.8, supported OpenMPI version to 4.0

CUDA implementation of TRG (flag cuda in solve_TRG()) based on Randomized-SVD.
TRG by four 3-legged tensors with matrix-free SVD (Morita et al) (MIZK_* flags in solve_TRG())
Added video tutorial for SuSMoST installation
Create samples with lattice vectors that are non collinear with surface unit cell lattice vectors
Support of joined states made by a couple of adsorption complexes in a single unit cell
ASE Espresso usage for lateral interactions computation
Added tutorials for Langmuir and binary gas mixture model with Monte Carlo and transfer-matrix methods
Added numerous models
Builds for multiple distribution: Centos7, Ubuntu Bionic, Debian Stable and Testing
Calculation of lattice state fingerprints for genetic algorithms

Transition to Python 3
Symmetry-based generation of lattice task from XYZ-files of adsorption complexes
Mean field and quasichemical approximation method added
Autocorrelation functions analysis added for Metropolis algorithm