|
Jive reference manual
|
Implements a direct solver that is based on sparse LU factorization with pivoting. More...
Public Types | |
| typedef algebra::SparseMatrixObject | MatrixObject |
| A type representing the coefficient matrix of a system of equations. More... | |
 Public Types inherited from jive::solver::Solver | |
| typedef jem::util::Properties | Properties |
| A type representing a set of properties. More... | |
| typedef util::Vector | Vector |
| A type that represents a vector. More... | |
Public Member Functions | |
| SparseLU (const jem::Ref< MatrixObject > &matrix) | |
Constructs a new SparseLU solver given a matrix object. More... | |
| void | factor () |
| Computes the LU factorization of the coefficient matrix. More... | |
| virtual void | solve (const Vector &lhs, const Vector &rhs) |
| Computes the solution of a linear system of equations. More... | |
| bool | enableAutoFactor (bool choice) |
| Sets the auto-factorization mode. More... | |
| virtual void | setZeroPivotThreshold (double eps) |
| Sets the threshold for detecting zero pivots. More... | |
| virtual double | getZeroPivotThreshold () const |
| Returns the current zero pivot threshold. More... | |
| virtual void | setMaxZeroPivotCount (int n) |
| Sets the maximum number of allowed zero pivots. More... | |
| virtual int | getMaxZeroPivotCount () const |
| Returns the maximum number of allowed zero pivots. More... | |
| Public Member Functions inherited from jive::solver::DirectSolver | |
| virtual void | configure (const Properties &conf, const Properties &props) |
| Configures this solver given a set of properties. More... | |
| Public Member Functions inherited from jive::solver::Solver | |
| virtual jem::Class * | getClass () const |
Returns the Class instance representing the runtime class of this object. More... | |
| Public Member Functions inherited from jem::Object | |
| virtual String | toString () const |
| Returns a short textual description of this object. More... | |
| virtual long | hashValue () const |
| Returns a hash value for this object. More... | |
| virtual bool | equals (const Ref< Object > &obj) const |
| Tests whether two objects are equal. More... | |
| Ref< Object > | clone () const |
| Returns a copy of this object. More... | |
Public Attributes | |
| jem::util::Event< int > | zeroPivotEvent |
| Signals the occurrence of a zero pivot. More... | |
| Public Attributes inherited from jive::solver::DirectSolver | |
| jem::util::Event< double > | progressEvent |
| Signals that progress is being made. More... | |
Protected Member Functions | |
| virtual | ~SparseLU () |
| Protected Member Functions inherited from jive::solver::DirectSolver | |
| virtual | ~DirectSolver () |
| Protected Member Functions inherited from jive::solver::Solver | |
| virtual | ~Solver () |
| Protected Member Functions inherited from jem::Collectable | |
| Collectable () | |
Creates an empty Collectable. More... | |
| ~Collectable () | |
| Frees resources. More... | |
Additional Inherited Members | |
| Static Public Member Functions inherited from jive::solver::Solver | |
| static jem::Class * | getType () |
| Static Public Member Functions inherited from jem::Object | |
| static Class * | getType () |
Returns the Class instance representing the Object class. More... | |
| Static Public Attributes inherited from jive::solver::DirectSolver | |
| static const double | DEFAULT_ZERO_PIVOT_THRESHOLD |
| The default zero pivot threshold. More... | |
#include<jive/solver/SparseLU.h>
The class SparseLU implements a direct solution algorithm that is based on LU factorization with partial (row-wise) pivoting. It can be used to solve any linear system of equations as long as the coefficient matrix is square and non-singular.
By default, a SparseLU instance will automatically update the LU factorization of the coefficient matrix whenever that is necessary. You can disable this behavior by calling the member function enableAutoFactor.
Although the SparseLU solver is often not the most efficient solver, it is one of the most robust solvers in Jive.
The MatrixObject type represents the coefficient matrix of a linear system of equations. It is an alias for jive::algebra::SparseMatrixObject.
|
explicit |
Constructs a new SparseLU solver that encapsulates a reference to the coefficient matrix matrix. The solver connects itself to the changeEvent of the matrix so that it receives an event whenever the matrix has been modified. In this way, the solver automatically knows when it is necessary to (re-)compute the LU factorization of the coefficient matrix.
| matrix | - the coefficient matrix of the linear system of equations to be solved. |
matrix != NIL
|
protectedvirtual |
| void jive::solver::SparseLU::factor | ( | ) |
Computes the LU factorization of the coefficient matrix that has been passed to the constructor of this SparseLU object. This function does nothing if the current LU factorization is up to date; that is, if the coefficient matrix has not been modified since the last time that this function was called.
If a zero pivot is encountered during the factorization procedure, and the maximum number of zero pivots has not yet been exceeded, the zero pivot is replaced by an infinitely large number. That is, the inverse of the pivot is set to zero. A pivot is marked as zero if its value divided by a scale factor is less than the zero pivot threshold. The scale factors are equal to the maximum row elements – in absolute value – of the coefficient matrix.
| SolverException | - if the maximum number of allowed zero pivots is exceeded. |
| jem::IllegalArgumentException | - if the coefficient matrix is non-square. |
Computes the solution of
L * U * lhs = rhs
with L and U the lower and upper factors of the coefficient matrix that has been passed to the constructor of this SparseLU object. This function calls the factor() function to (re-)compute the lower and upper factors, unless the auto-factorization mode has been disabled by the enableAutoFactor() function.
When this function is called, the vector rhs should contain the right-hand side vector. On exit, the vector lhs will be set to the solution of the system of equations. The sizes of the two vectors must match the size of the coefficient matrix.
| lhs | - the solution vector. |
| rhs | - the right-hand side vector. |
| SolverException | - if the maximum number of allowed zero pivots is exceeded. |
| jem::IllegalArgumentException | - if the coefficient matrix is non-square. |
Implements jive::solver::Solver.
| bool jive::solver::SparseLU::enableAutoFactor | ( | bool | choice | ) |
Enables (when choice is true) or disables (when choice is false) the auto-factorization mode. If this mode is enabled, the function factor() is automatically called by the solve() function. Otherwise, you are responsible for calling the factor() function.
| choice | - a boolean value that indicates whether the auto-factorization mode should be enabled or not. |
true if the auto-factorization mode was enabled before this function was called, and false otherwise.
|
virtual |
Sets the zero pivot threshold to eps.
| eps | - the relative zero pivot threshold. |
eps > 0.0this->getZeroPivotThreshold() == eps Implements jive::solver::DirectSolver.
|
virtual |
Returns the current zero pivot threshold.
Implements jive::solver::DirectSolver.
|
virtual |
Sets the maximum number of allowed zero pivots to n.
| n | - the maximum number of allowed zero pivots. |
n >= 0this->getMaxZeroPivotCount() == n Implements jive::solver::DirectSolver.
|
virtual |
Returns the maximum number of allowed zero pivots.
Implements jive::solver::DirectSolver.
| jem::util::Event<int> jive::solver::SparseLU::zeroPivotEvent |
The event zeroPivotEvent is emitted whenever a zero pivot is encountered. The event argument is the index of the matrix column containing the zero pivot.
