trilinos Package

trilinos Package

fipy.solvers.trilinos.DefaultSolver

alias of LinearGMRESSolver

fipy.solvers.trilinos.DummySolver

alias of LinearGMRESSolver

fipy.solvers.trilinos.DefaultAsymmetricSolver

alias of LinearGMRESSolver

fipy.solvers.trilinos.GeneralSolver

alias of LinearGMRESSolver

class fipy.solvers.trilinos.LinearCGSSolver(tolerance=1e-10, iterations=1000, precon=<fipy.solvers.trilinos.preconditioners.multilevelDDPreconditioner.MultilevelDDPreconditioner instance at 0x111354098>)

Bases: fipy.solvers.trilinos.trilinosAztecOOSolver.TrilinosAztecOOSolver

The LinearCGSSolver is an interface to the cgs solver in Trilinos, using the MultilevelSGSPreconditioner by default.

Parameters :

• tolerance: The required error tolerance.
• iterations: The maximum number of iterative steps to perform.
• precon: Preconditioner to use.

class fipy.solvers.trilinos.LinearPCGSolver(tolerance=1e-10, iterations=1000, precon=<fipy.solvers.trilinos.preconditioners.multilevelDDPreconditioner.MultilevelDDPreconditioner instance at 0x11134a908>)

Bases: fipy.solvers.trilinos.trilinosAztecOOSolver.TrilinosAztecOOSolver

The LinearPCGSolver is an interface to the cg solver in Trilinos, using the MultilevelSGSPreconditioner by default.

Parameters :

• tolerance: The required error tolerance.
• iterations: The maximum number of iterative steps to perform.
• precon: Preconditioner to use.

class fipy.solvers.trilinos.LinearGMRESSolver(tolerance=1e-10, iterations=1000, precon=<fipy.solvers.trilinos.preconditioners.multilevelDDPreconditioner.MultilevelDDPreconditioner instance at 0x1113540e0>)

Bases: fipy.solvers.trilinos.trilinosAztecOOSolver.TrilinosAztecOOSolver

The LinearGMRESSolver is an interface to the gmres solver in Trilinos, using a the MultilevelDDPreconditioner by default.

Parameters :

• tolerance: The required error tolerance.
• iterations: The maximum number of iterative steps to perform.
• precon: Preconditioner to use.

class fipy.solvers.trilinos.LinearLUSolver(tolerance=1e-10, iterations=10, precon=None, maxIterations=10)

Bases: fipy.solvers.trilinos.trilinosSolver.TrilinosSolver

The LinearLUSolver is an interface to the Amesos KLU solver in Trilinos.

Parameters :

• tolerance: The required error tolerance.
• iterations: The maximum number of iterative steps to perform.

class fipy.solvers.trilinos.LinearBicgstabSolver(tolerance=1e-10, iterations=1000, precon=<fipy.solvers.trilinos.preconditioners.jacobiPreconditioner.JacobiPreconditioner instance at 0x1113541b8>)

Bases: fipy.solvers.trilinos.trilinosAztecOOSolver.TrilinosAztecOOSolver

The LinearBicgstabSolver is an interface to the biconjugate gradient stabilized solver in Trilinos, using the JacobiPreconditioner by default.

Parameters :

• tolerance: The required error tolerance.
• iterations: The maximum number of iterative steps to perform.
• precon: Preconditioner to use.

class fipy.solvers.trilinos.TrilinosMLTest(tolerance=1e-10, iterations=5, MLOptions={}, testUnsupported=False)

Bases: fipy.solvers.trilinos.trilinosSolver.TrilinosSolver

This solver class does not actually solve the system, but outputs information about what ML preconditioner settings will work best.

Parameters :

• tolerance: The required error tolerance.
• iterations: The maximum number of iterations to perform per test.
• MLOptions: Options to pass to ML. A dictionary of {option:value} pairs. This will be passed to ML.SetParameterList.
• testUnsupported: test smoothers that are not currently implemented in preconditioner objects.

For detailed information on the possible parameters for ML, see http://trilinos.sandia.gov/packages/ml/documentation.html

Currently, passing options to Aztec through ML is not supported.

class fipy.solvers.trilinos.MultilevelDDPreconditioner

Bases: fipy.solvers.trilinos.preconditioners.preconditioner.Preconditioner

Multilevel preconditioner for Trilinos solvers. A classical smoothed aggregation-based 2-level domain decomposition.

Create a Preconditioner object.

class fipy.solvers.trilinos.MultilevelSAPreconditioner

Bases: fipy.solvers.trilinos.preconditioners.preconditioner.Preconditioner

Multilevel preconditioner for Trilinos solvers suitable classical smoothed aggregation for symmetric positive definite or nearly symmetric positive definite systems.

Create a Preconditioner object.

class fipy.solvers.trilinos.MultilevelDDMLPreconditioner

Bases: fipy.solvers.trilinos.preconditioners.preconditioner.Preconditioner

Multilevel preconditioner for Trilinos solvers. 3-level algebraic domain decomposition.

Create a Preconditioner object.

class fipy.solvers.trilinos.MultilevelNSSAPreconditioner

Bases: fipy.solvers.trilinos.preconditioners.preconditioner.Preconditioner

Energy-based minimizing smoothed aggregation suitable for highly convective non-symmetric fluid flow problems.

Create a Preconditioner object.

class fipy.solvers.trilinos.JacobiPreconditioner

Bases: fipy.solvers.trilinos.preconditioners.preconditioner.Preconditioner

Jacobi Preconditioner for Trilinos solvers.

Create a Preconditioner object.

class fipy.solvers.trilinos.ICPreconditioner

Bases: fipy.solvers.trilinos.preconditioners.preconditioner.Preconditioner

Incomplete Cholesky Preconditioner from IFPACK for Trilinos Solvers.

Create a Preconditioner object.

class fipy.solvers.trilinos.DomDecompPreconditioner

Bases: fipy.solvers.trilinos.preconditioners.preconditioner.Preconditioner

Domain Decomposition preconditioner for Trilinos solvers.

Create a Preconditioner object.

class fipy.solvers.trilinos.MultilevelSGSPreconditioner(levels=10)

Bases: fipy.solvers.trilinos.preconditioners.preconditioner.Preconditioner

Multilevel preconditioner for Trilinos solvers using Symmetric Gauss-Seidel smoothing.

Initialize the multilevel preconditioner

• levels: Maximum number of levels

class fipy.solvers.trilinos.MultilevelSolverSmootherPreconditioner(levels=10)

Bases: fipy.solvers.trilinos.preconditioners.preconditioner.Preconditioner

Multilevel preconditioner for Trilinos solvers using Aztec solvers as smoothers.

Initialize the multilevel preconditioner

• levels: Maximum number of levels

linearBicgstabSolver Module

class fipy.solvers.trilinos.linearBicgstabSolver.LinearBicgstabSolver(tolerance=1e-10, iterations=1000, precon=<fipy.solvers.trilinos.preconditioners.jacobiPreconditioner.JacobiPreconditioner instance at 0x1113541b8>)

Bases: fipy.solvers.trilinos.trilinosAztecOOSolver.TrilinosAztecOOSolver

The LinearBicgstabSolver is an interface to the biconjugate gradient stabilized solver in Trilinos, using the JacobiPreconditioner by default.

Parameters :

• tolerance: The required error tolerance.
• iterations: The maximum number of iterative steps to perform.
• precon: Preconditioner to use.

linearCGSSolver Module

class fipy.solvers.trilinos.linearCGSSolver.LinearCGSSolver(tolerance=1e-10, iterations=1000, precon=<fipy.solvers.trilinos.preconditioners.multilevelDDPreconditioner.MultilevelDDPreconditioner instance at 0x111354098>)

Bases: fipy.solvers.trilinos.trilinosAztecOOSolver.TrilinosAztecOOSolver

The LinearCGSSolver is an interface to the cgs solver in Trilinos, using the MultilevelSGSPreconditioner by default.

Parameters :

• tolerance: The required error tolerance.
• iterations: The maximum number of iterative steps to perform.
• precon: Preconditioner to use.

linearGMRESSolver Module

class fipy.solvers.trilinos.linearGMRESSolver.LinearGMRESSolver(tolerance=1e-10, iterations=1000, precon=<fipy.solvers.trilinos.preconditioners.multilevelDDPreconditioner.MultilevelDDPreconditioner instance at 0x1113540e0>)

Bases: fipy.solvers.trilinos.trilinosAztecOOSolver.TrilinosAztecOOSolver

The LinearGMRESSolver is an interface to the gmres solver in Trilinos, using a the MultilevelDDPreconditioner by default.

Parameters :

• tolerance: The required error tolerance.
• iterations: The maximum number of iterative steps to perform.
• precon: Preconditioner to use.

linearLUSolver Module

class fipy.solvers.trilinos.linearLUSolver.LinearLUSolver(tolerance=1e-10, iterations=10, precon=None, maxIterations=10)

Bases: fipy.solvers.trilinos.trilinosSolver.TrilinosSolver

The LinearLUSolver is an interface to the Amesos KLU solver in Trilinos.

Parameters :

• tolerance: The required error tolerance.
• iterations: The maximum number of iterative steps to perform.

linearPCGSolver Module

class fipy.solvers.trilinos.linearPCGSolver.LinearPCGSolver(tolerance=1e-10, iterations=1000, precon=<fipy.solvers.trilinos.preconditioners.multilevelDDPreconditioner.MultilevelDDPreconditioner instance at 0x11134a908>)

Bases: fipy.solvers.trilinos.trilinosAztecOOSolver.TrilinosAztecOOSolver

The LinearPCGSolver is an interface to the cg solver in Trilinos, using the MultilevelSGSPreconditioner by default.

Parameters :

• tolerance: The required error tolerance.
• iterations: The maximum number of iterative steps to perform.
• precon: Preconditioner to use.

trilinosAztecOOSolver Module

class fipy.solvers.trilinos.trilinosAztecOOSolver.TrilinosAztecOOSolver(tolerance=1e-10, iterations=1000, precon=<fipy.solvers.trilinos.preconditioners.jacobiPreconditioner.JacobiPreconditioner instance at 0x111354050>)

Bases: fipy.solvers.trilinos.trilinosSolver.TrilinosSolver

Attention

This class is abstract, always create on of its subclasses. It provides the code to call all solvers from the Trilinos AztecOO package.

Parameters :

• tolerance: The required error tolerance.
• iterations: The maximum number of iterative steps to perform.
• precon: Preconditioner object to use.

trilinosMLTest Module

class fipy.solvers.trilinos.trilinosMLTest.TrilinosMLTest(tolerance=1e-10, iterations=5, MLOptions={}, testUnsupported=False)

Bases: fipy.solvers.trilinos.trilinosSolver.TrilinosSolver

This solver class does not actually solve the system, but outputs information about what ML preconditioner settings will work best.

Parameters :

• tolerance: The required error tolerance.
• iterations: The maximum number of iterations to perform per test.
• MLOptions: Options to pass to ML. A dictionary of {option:value} pairs. This will be passed to ML.SetParameterList.
• testUnsupported: test smoothers that are not currently implemented in preconditioner objects.

For detailed information on the possible parameters for ML, see http://trilinos.sandia.gov/packages/ml/documentation.html

Currently, passing options to Aztec through ML is not supported.

trilinosNonlinearSolver Module

class fipy.solvers.trilinos.trilinosNonlinearSolver.TrilinosNonlinearSolver(equation, jacobian=None, tolerance=1e-10, iterations=1000, printingOptions=None, solverOptions=None, linearSolverOptions=None, lineSearchOptions=None, directionOptions=None, newtonOptions=None)

Bases: fipy.solvers.trilinos.trilinosSolver.TrilinosSolver

solve(dt=None)

trilinosSolver Module

class fipy.solvers.trilinos.trilinosSolver.TrilinosSolver(*args, **kwargs)

Bases: fipy.solvers.solver.Solver

Attention

This class is abstract. Always create one of its subclasses.

Subpackages

• preconditioners Package
  • preconditioners Package
  • domDecompPreconditioner Module
  • icPreconditioner Module
  • jacobiPreconditioner Module
  • multilevelDDMLPreconditioner Module
  • multilevelDDPreconditioner Module
  • multilevelNSSAPreconditioner Module
  • multilevelSAPreconditioner Module
  • multilevelSGSPreconditioner Module
  • multilevelSolverSmootherPreconditioner Module
  • preconditioner Module