Vandermonde systems on equidistant nodes in \([0,1]\): accurate computation (Q2489396): Difference between revisions
From MaRDI portal
Changed an Item |
Changed an Item |
||
| Property / describes a project that uses | |||
| Property / describes a project that uses: mctoolbox / rank | |||
Normal rank | |||
Revision as of 10:41, 29 February 2024
scientific article
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | Vandermonde systems on equidistant nodes in \([0,1]\): accurate computation |
scientific article |
Statements
Vandermonde systems on equidistant nodes in \([0,1]\): accurate computation (English)
0 references
28 April 2006
0 references
Arguments from integer number theory are utilized to determine an explicit factorization for the entries of the inverse of a Vandermonde matrix with equidistant nodes. Some recursions for the practical implementation of the proposed formula are given. The authors also give formulas for the Frobenius norms of the matrix and its inverse. Simple asymptotic estimates of these norms are given. Results of numerical experiments showing an accurate representation of the floating point number system are reported.
0 references
Vandermonde matrices
0 references
polynomial interpolation
0 references
conditioning
0 references
inverse
0 references
equidistant nodes
0 references
Frobenius norms
0 references
numerical experiments
0 references
floating point number system
0 references
