Partitioned multiply scaled pseudo conjugate gradient schemes

From MaRDI portal





The authors consider the system of linear algebraic equations \(Ax=b\) with a symmetric positive definite matrix \(A\). They suppose that the matrix \(A\) has a block form with \(p\) diagonal square blocks. The authors develop the classical algorithm of the conjugate gradients, where in every step two parameters are calculated in the sense that two diagonal matrices, which is analogous to the two parameters in the classical case, are calculated in each step. These block diagonal matrices have the same block form as the matrix \(A\). Numerical tests for the heat conduction problem are presented for preconditioned and nonpreconditioned versions and various partitions of the matrix \(A\).











This page was built for publication: Partitioned multiply scaled pseudo conjugate gradient schemes

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1185504)