Go to the source code of this file.
|
| namespace | ug |
| | the ug namespace
|
| |
|
| template<typename Matrix_type , typename Vector_type > |
| void | ug::diag_step (const Matrix_type &A, Vector_type &c, const Vector_type &d, number damp) |
| | Performs a jacobi-step.
|
| |
| template<typename Matrix_type , typename Vector_type > |
| void | ug::gs_step_LL (const Matrix_type &A, Vector_type &c, const Vector_type &d, const number relaxFactor) |
| | Gauss-Seidel-Iterations.
|
| |
| template<typename Matrix_type , typename Vector_type > |
| void | ug::gs_step_UR (const Matrix_type &A, Vector_type &c, const Vector_type &d, const number relaxFactor) |
| | Performs a backward gauss-seidel-step, that is, solve on the upper right of A. Using gs_step_UR within a preconditioner-scheme leads to the fact that we get the correction by successive inserting the already computed values of c in c = N * d, with c being the correction and d being the defect. N denotes the matrix of the second normal-form of a linear iteration scheme.
|
| |
| template<typename Matrix_type , typename Vector_type > |
| void | ug::sgs_step (const Matrix_type &A, Vector_type &c, const Vector_type &d, const number relaxFactor) |
| | Performs a symmetric gauss-seidel step. Using sgs_step within a preconditioner-scheme leads to the fact that we get the correction by successive inserting the already computed values of c in c = N * d, with c being the correction and d being the defect. N denotes the matrix of the second normal-form of a linear iteration scheme.
|
| |