Package maspack.solvers

Interface Summary

Interface	Description
DirectSolver
IterativeSolver
LCPSolver	General interface for linear complementarity problem (LCP) solvers.

Class Summary

Class	Description
AccelerateSolver	A direct solver based on the sparse factorizations in Apple's Accelerate framework (Sparse/Solve.h), for comparison with MUMPS on Arm-based Macs, where Pardiso is not available.
AccelerateSolverTest	Test for AccelerateSolver.
CGSolver	Solves linear systems using the conjugate gradient algorithm
CGSolverTest	Unit test for the class CGSolver
CRSolver
CRSolverTest	Unit test for the class CGSolver
DantzigLCPSolver	Solves linear complementarity problems (LCPs) and bounded linear complementarity problems (BLCPs) for symmetric positive semi-definite (SPSD) matrices using a principal pivoting method.
DantzigLCPSolverTest
DantzigQPSolver	A dense QP (Quadratic Program) solver that that uses Dantzig's algorithm.
DantzigQPSolverTest	Testing program for the QPSolver
DirectSolverBase	Base class for DirectSolver implementations which are implemented natively and which accept their matrices in compressed row storage (CRS) format with 1-based indices.
DirectSolverTestBase	Base class for testing DirectSolver implementations.
DirectSolverTiming	Times the analyze, factor and solve phases of the direct sparse solvers across a range of matrix sizes, using the synthetic 2D and 3D finite element matrices produced by FemMatrixGenerator.
FemMatrixGenerator	Generates synthetic sparse matrices whose sparsity patterns match those arising from 2D and 3D finite element models on structured grids.
FemMatrixGenerator.CRSMatrix	Sparse symmetric matrix in the 1-based upper triangular CRS form used by DirectSolver.
FindSolverThreadThreshold	Finds the single matrix-size threshold below which a DirectSolverBase subclass should be forced to one thread, and prints the resulting ThreadLimit table entry -- see DirectSolverBase.maxThreadsNnz(int, int), DirectSolverBase.ThreadLimit and DirectSolverBase.lookupMaxThreads(maspack.solvers.DirectSolverBase.ThreadLimit[], int) -- ready to paste into that solver's override.
HybridReplayTiming	Replays the solves recorded by MurtyMechSolver.writeSolveInfo(java.io.PrintWriter, maspack.solvers.LCPSolver.Status, maspack.matrix.VectorNd, maspack.matrix.VectorNd, maspack.matrix.VectorNd, maspack.matrix.VectorNd, maspack.matrix.VectorNd, maspack.matrix.VectorNd, maspack.matrix.SparseBlockMatrix, int, maspack.matrix.VectorNd, int, maspack.matrix.SparseBlockMatrix, maspack.matrix.VectorNd, maspack.matrix.VectorNd, maspack.matrix.SparseBlockMatrix, maspack.matrix.VectorNd, maspack.matrix.VectorNd, maspack.matrix.VectorNi, maspack.matrix.SparseBlockMatrix, maspack.matrix.VectorNd, maspack.matrix.VectorNd, maspack.matrix.VectorNi, java.util.ArrayList<maspack.spatialmotion.FrictionInfo>, int, int) for a simulation with (at most) bilateral constraints, to measure hybrid direct/iterative solves on real ArtiSynth matrix sequences.
HybridSolvePolicy	Decides when hybrid direct/iterative solves should use an iterative solve (preconditioned by the most recent factorization) and when they should refactor.
HybridSolvePolicyTest	Tests the refactor and backoff decisions of HybridSolvePolicy using scripted solve times.
HybridSolveTiming	Measures the performance of hybrid direct/iterative solves, in which a matrix factored for values V0 is used as a preconditioner for an iterative solve with nearby values V1, as done by DirectSolver.iterativeSolve(double[], double[], double[]).
IncompleteCholeskyDecomposition	Constructs the Cholesky decomposition of a symmetric positive definite matrix.
IncompleteCholeskyDecompositionTest
IncompleteLUDecomposition	Constructs the Cholesky decomposition of a symmetric positive definite matrix.
IncompleteLUDecompositionTest
KKTSolver
KKTSolverTest
LCPSolverTestBase	Base class for LCP solver tests
LemkeContactSolver
LemkeContactSolverTest
LemkeSolverBase
MumpsSolver	JNI interface to the MUMPS sparse solver.
MumpsSolverTest	Unit test for MumpsSolver.
MurtyLCPSolver	Solves linear complementarity problems (LCPs) and bounded linear complementarity problems (BLCPs) for symmetric positive definite (SPD) matrices using Murty's method.
MurtyLCPSolverTest
MurtyMechSolver	Solves a bounded mixed linear complementarity problem (MLCP) for a system arising from velocity solves for mechanical systems including bilateral constraints (e.g.
MurtyMechSolverTest
NMFactorization	Computes a non-negative matrix factorization (NMF) of a non-negative m X n matrix V:
NMFactorizationTest
PardisoSolver	JNI interface to the Pardiso sparse solver.
PardisoSolverTest
PardisoThreadTest
PardisoThreadTest.SparseElement
PhysicalCoreCountTiming	Quick timing check for DirectSolverBase.getPhysicalCoreCountNative(), backed by hybridGetPhysicalCoreCount() in hybridSolve.cc.
ProjectedCGSolver	Solves linear systems using the conjugate gradient algorithm
SimpleLCPSolver	Solves linear complementarity problems (LCPs) for symmetric positive semi-definite (SPSD) matrices using the simple principal pivoting method.
SimpleLCPSolverTest
TriDiagonalSolver	Solvers for tridiagonal matrices, using the Thomas algorithms as described in the wikipedia entry for "Tridiagonal matrix algorithm".
TriDiagonalSolverTest	Test class for TriDiagonalSolver.
UmfpackSolver
UmfpackSolverTest

Enum Summary

Enum	Description
DantzigQPSolver.Status	Described whether or not a solution was found.
DirectSolver.IterativeMethod	Describes the methods available for DirectSolver.iterativeSolve(double[], double[], double[]), each of which uses the most recent factorization as a preconditioner.
FemMatrixGenerator.ElemType	Element type used to subdivide the grid.
IterativeSolver.ToleranceType
KKTSolver.Status	Described whether or not a solution was found.
LCPSolver.Status	Solver return status
MumpsSolver.ReorderMethod	Describes the reorder methods that can be used during the analyze phase to reduce factorization fill-in.
NMFactorization.Init	Method used to initialize the factors.
NMFactorization.Normalization	Normalization applied to the factors once the factorization is complete.
NMFactorization.Update	Method used to update the factors at each iteration.
PardisoSolver.ReorderMethod	Describes the reorder methods that can be used during the analyze phase to reduce factorization fill-in.
SparseSolverId	Identifies the general-purpose sparse solvers available in this package.