petsc_ksp.py

LinearSolver that uses PetSC KSP to solve for a system’s derivatives.

class openmdao.solvers.linear.petsc_ksp.Monitor(solver)

Bases: object

Prints output from PETSc’s KSP solvers.

Callable object given to KSP as a callback for printing the residual.

Parameters:

solverobject

The openmdao solver.

Attributes:

_solver_solver

The openmdao solver.

_normfloat

The current norm.

_norm0float

The norm for the first iteration.

__init__(solver)

Store pointer to the openmdao solver and initialize norms.

class openmdao.solvers.linear.petsc_ksp.PETScKrylov(**kwargs)

Bases: LinearSolver

LinearSolver that uses PetSC KSP to solve for a system’s derivatives.

Parameters:

**kwargsdict

Dictionary of options set by the instantiating class/script.

Attributes:

preconSolver

Preconditioner for linear solve. Default is None for no preconditioner.

_kspdist

Dictionary of KSP instances (keyed on vector name).

_lin_rhs_checkerLinearRHSChecker or None

Object for checking the right-hand side of the linear solve.

SOLVER = 'LN: PETScKrylov'

__init__(**kwargs)

Declare the solver options.

add_recorder(recorder)

Add a recorder to the solver’s RecordingManager.

Parameters:

recorder<CaseRecorder>

A recorder instance to be added to RecManager.

apply(mat, in_vec, result)

Apply preconditioner.

Parameters:

matPETSc.Mat

PETSc matrix object.

in_vecPETSc.Vector

Incoming vector.

resultPETSc.Vector

Empty vector in which the preconditioned in_vec is stored.

can_solve_cycle()

Return True if this solver can solve groups with cycles.

Returns:

bool

True if this solver can solve groups with cycles.

cleanup()

Clean up resources prior to exit.

does_recursive_applies()

Return False.

By default, assume linear solvers don’t do recursive apply_linear calls.

Returns:

bool

True if solver makes recursive apply_linear calls on its subsystems.

get_outputs_dir(*subdirs, mkdir=True)

Get the path under which all output files of this solver are to be placed.

Parameters:

*subdirsstr

Subdirectories nested under the relevant problem output directory. To create {prob_output_dir}/a/b one would pass solver.get_outputs_dir(‘a’, ‘b’).

mkdirbool

If True, attempt to create this directory if it does not exist.

Returns:

pathlib.Path

The path of the outputs directory for the problem.

get_reports_dir()

Get the path to the directory where the report files should go.

If it doesn’t exist, it will be created.

Returns:

str

The path to the directory where reports should be written.

property msginfo

Return info to prepend to messages.

Returns:

str

Info to prepend to messages.

mult(mat, in_vec, result)

Apply Jacobian matrix (KSP Callback).

The following attributes must be defined when solve is called to provide information used in this callback:

_systemSystem

pointer to the owning system.

_vec_namestr

the right-hand-side (RHS) vector name.

_modestr

‘fwd’ or ‘rev’.

Parameters:

matPETSc.Mat

PETSc matrix object.

in_vecPetSC Vector

Incoming vector.

resultPetSC Vector

Empty array into which we place the matrix-vector product.

record_iteration(**kwargs)

Record an iteration of the current Solver.

Parameters:

**kwargsdict

Keyword arguments (used for abs and rel error).

report_failure(msg)

Report a failure that has occurred.

The msg may be printed or ignored depending on the iprint option, and may raise an AnalysisError depending on the err_on_non_converge option.

Parameters:

msgstr

Message indicating the failure.

solve(mode, rel_systems=None)

Solve the linear system for the problem in self._system().

The full solution vector is returned.

Parameters:

modestr

Derivative mode, can be ‘fwd’ or ‘rev’.

rel_systemsset of str

Names of systems relevant to the current solve. Deprecated.

use_relevance()

Return True if relevance should be active.

Returns:

bool

True if relevance should be active.