General Options
The general options are placed inside an Options block:
begin Options
...
endHamster.get_nhamster — Method
nhamster=1The nhamster tag sets the number of Hamster processes to be spawned for parallel tasks.
Hamster.get_nthreads_bands — Method
nthreads_bands=JULIANUMTHREADS
The nthreads_bands::Int tag sets the number of tasks to work on energy bands simultaneously.
Hamster.get_nthreads_blas — Method
nthreads_blas=1
The nthreads_blas::Int tag sets the number of threads used by the BLAS library.
Hamster.get_nthreads_kpoints — Method
nthreads_kpoints=JULIANUMTHREADS
The nthreads_kpoints::Int tag sets the number of tasks to work on kpoints simultaneously.
Hamster.set_seed! — Method
seed=noneThe seed tag can be used to set a custom seed for RNG.
Hamster.get_orbitals — Method
orbitals=[]
The orbitals tag specifies the set of orbitals used as the basis for the ion of the given type. If left empty, no orbitals are assigned and the corresponding atomic species is excluded from the electronic structure.
Hamster.get_rllm_file — Method
rllm_file=rllm.dat
The rllm_file::String tag sets the name of the file where the distance dependence is stored.
Hamster.get_tmethod — Method
tmethod=rotation
The tmethod::String tag selects the method used to compute the Slater-Koster transformation matrix. In practice, this choice should not affect the results, although rotation is slightly more efficient.
Possible options:
rotation: constructs the new reference system from the rotation matrix (default).gramschmidt: constructs the new reference system using a Gram-Schmidt procedure.
Hamster.SparsityMode — Type
Abstract type SparsityMode serves as a base for indicating the sparsity of matrices. These types enable dispatching based on the sparsity mode.
Subtypes
Sparse <: SparsityMode: Represents matrices with sparse storage.Dense <: SparsityMode: Represents matrices with dense storage.
Hamster.get_sp_diag — Method
sp_diag=false
The sp_diag::Bool tag switches between dense and sparse methods for matrix diagonalization (only affects Hᵏ and diagonalization). This can not be combined with optimization.
Hamster.get_sp_mode — Method
sp_mode=false
The sp_mode::Bool tag switches between dense and sparse matrix methods. This will only affect the computation of Hᴿ (not Hᵏ or diagonalization) and gradient computations when doing optimization.