Spectral Differentiation: Integration and Inversion

16.10.2018 14:00

Pseudospectral differentiation matrices have large round-off error and can lead to ill-conditioned systems used to solve differential equations numerically. This talk presents two matrices designed to precondition these systems and improve robustness. The first is a generalization of a pseudospectral integration matrix and the second makes use of this matrix to construct the matrix representing the inverse operator of the differential equation. We examine expected and computed eigenvalues of some of the matrices involved. We test the preconditioners on a variety of examples.


salle 623, Séminaire d'analyse numérique

Organisé par

Section de mathématiques


Conor McCoid, Université de Genève

entrée libre


Catégorie: Séminaire

Mots clés: analyse numérique