gemseo / algos / linear_solvers

lib_scipy_linalg module¶

Wrappers for SciPy’s linear solvers.

class gemseo.algos.linear_solvers.lib_scipy_linalg.ScipyLinalgAlgos[source]¶

Bases: LinearSolverLibrary

Wrapper for scipy linalg sparse linear solvers.

driver_has_option(option_name)¶

Check the existence of an option.

Parameters:: option_name (str) – The name of the option.
Returns:: Whether the option exists.
Return type:: bool

execute(problem, algo_name=None, **options)¶

Execute the driver.

Parameters:

problem (BaseProblem) – The problem to be solved.
algo_name (str | None) – The name of the algorithm. If None, use algo_name attribute which may have been set by the factory.
**options (Any) – The algorithm options.

Return type:

None

filter_adapted_algorithms(problem)¶

Filter the algorithms capable of solving the problem.

Parameters:: problem (Any) – The problem to be solved.
Returns:: The names of the algorithms adapted to this problem.
Return type:: list[str]

classmethod get_default_properties(algo_name)[source]¶

Return the properties of the algorithm.

It states if it requires symmetric, or positive definite matrices for instance.

Parameters:: algo_name (str) – The algorithm name.
Returns:: The properties of the solver.
Return type:: LinearSolverDescription

init_options_grammar(algo_name)¶

Initialize the options’ grammar.

Parameters:: algo_name (str) – The name of the algorithm.
Return type:: JSONGrammar

classmethod is_algorithm_suited(algorithm_description, problem)¶

Check if an algorithm is suited to a problem according to its description.

Parameters:

algorithm_description (AlgorithmDescription) – The description of the algorithm.
problem (Any) – The problem to be solved.

Returns:

Whether the algorithm is suited to the problem.

Return type:

bool

solve(linear_problem, algo_name, **options)¶

Solve the linear system.

Parameters:

linear_problem (LinearProblem) – The problem to solve.
algo_name (str) – The name of the algorithm.
**options (Any) – The algorithm options.

Returns:

The execution result.

Return type:

Any

BASE_INFO_MSG = 'scipy linear solver algorithm stop info: '¶

LGMRES_SPEC_OPTS = ('inner_m', 'outer_k', 'outer_v', 'store_outer_av', 'prepend_outer_v')¶

LIBRARY_NAME: ClassVar[str | None] = 'SciPy'¶: The name of the interfaced library.

OPTIONS_DIR: ClassVar[str | Path] = 'options'¶: The name of the directory containing the files of the grammars of the options.

OPTIONS_MAP: ClassVar[dict[str, str]] = {'max_iter': 'maxiter', 'preconditioner': 'M', 'store_outer_av': 'store_outer_Av'}¶: The names of the options in GEMSEO mapping to those in the wrapped library.

SAVE_WHEN_FAIL = 'save_when_fail'¶

algo_name: str | None¶: The name of the algorithm used currently.

property algorithms: list[str]¶: The available algorithms.

descriptions: dict[str, AlgorithmDescription]¶: The description of the algorithms contained in the library.

internal_algo_name: str | None¶

The internal name of the algorithm used currently.

It typically corresponds to the name of the algorithm in the wrapped library if any.

methods_map: dict[str, Callable[[ndarray, ndarray, ...], tuple[ndarray, int]]]¶: The mapping between the solver names and the solvers methods in scipy.sparse.

opt_grammar: JSONGrammar | None¶: The grammar defining the options of the current algorithm.

problem: Any | None¶: The problem to be solved.

save_fpath: str¶: The path to the file to saved the problem when it is not converged and the option save_when_fail is active.

property solution: ndarray¶: The solution of the problem.