We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

math.FA
< prev | next >
new | recent | 1712

Change to browse by:

math
math.NA

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo

Mathematics > Numerical Analysis

Title: Matrix methods for Padé approximation: numerical calculation of poles, zeros and residues

Authors: Luca Perotti, Michal Wojtylak
(Submitted on 4 Dec 2017 (v1), last revised 17 Jan 2018 (this version, v2))
Abstract: A representation of the Pad\'e approximation of the $Z$-transform of a signal as a resolvent of a tridiagonal matrix $J$ is given. Several formulas for the poles, zeros and residues of the Pad\'e approximation in terms of the matrix $J$ are proposed. Their numerical stability is tested and compared. Methods for computing forward and backward errors are presented.
Subjects: Numerical Analysis (math.NA); Functional Analysis (math.FA)
Cite as: arXiv:1712.01155 [math.NA]
  (or arXiv:1712.01155v2 [math.NA] for this version)

Submission history

From: Michał Wojtylak [view email]
[v1] Mon, 4 Dec 2017 15:46:35 GMT (243kb,D)
[v2] Wed, 17 Jan 2018 15:29:44 GMT (245kb,D)
Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Link back to: arXiv, form interface, contact.