Namespaces
Variants
Actions

Difference between revisions of "Szegö quadrature"

From Encyclopedia of Mathematics
Jump to: navigation, search
m (moved Szegö quadrature to Szego quadrature: ascii title)
m (AUTOMATIC EDIT (latexlist): Replaced 20 formulas out of 20 by TEX code with an average confidence of 2.0 and a minimal confidence of 2.0.)
 
(One intermediate revision by one other user not shown)
Line 1: Line 1:
Szegö quadrature formulas are the analogues on the unit circle <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s1306601.png" /> in the complex plane of the Gauss quadrature formulas on an interval (cf. also [[Gauss quadrature formula|Gauss quadrature formula]]). They approximate the integral
+
<!--This article has been texified automatically. Since there was no Nroff source code for this article,
 +
the semi-automatic procedure described at https://encyclopediaofmath.org/wiki/User:Maximilian_Janisch/latexlist
 +
was used.
 +
If the TeX and formula formatting is correct, please remove this message and the {{TEX|semi-auto}} category.
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s1306602.png" /></td> </tr></table>
+
Out of 20 formulas, 20 were replaced by TEX code.-->
  
where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s1306603.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s1306604.png" /> is a positive [[Measure|measure]] on <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s1306605.png" />, by a [[Quadrature formula|quadrature formula]] of the form
+
{{TEX|semi-auto}}{{TEX|done}}
 +
Szegö quadrature formulas are the analogues on the unit circle $\bf T$ in the complex plane of the Gauss quadrature formulas on an interval (cf. also [[Gauss quadrature formula|Gauss quadrature formula]]). They approximate the integral
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s1306606.png" /></td> </tr></table>
+
\begin{equation*} I _ { \mu } ( f ) = \int _ { T } f ( t ) d \mu ( t ), \end{equation*}
  
One cannot take the zeros of the Szegö polynomials <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s1306607.png" /> as nodes (as in Gaussian formulas), because these are all in the open unit disc <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s1306608.png" /> (cf. also [[Szegö polynomial|Szegö polynomial]]). Therefore, the para-orthogonal polynomials are introduced as <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s1306609.png" />, where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s13066010.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s13066011.png" />. These are orthogonal to <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s13066012.png" /> and have <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s13066013.png" /> simple zeros, which are on <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s13066014.png" />. The Szegö quadrature formula then takes as nodes the zeros <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s13066015.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s13066016.png" />, of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s13066017.png" />, and as weights the [[Christoffel numbers|Christoffel numbers]]
+
where $\mathbf{T} = \{ z \in \mathbf{C} : | z | = 1 \}$ and $\mu$ is a positive [[Measure|measure]] on $\bf T$, by a [[Quadrature formula|quadrature formula]] of the form
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s13066018.png" /></td> </tr></table>
+
\begin{equation*} I _ { n } ( f ) = \sum _ { k = 1 } ^ { n } \lambda _ { n k } f ( \xi _ { n k } ). \end{equation*}
  
The result is a quadrature formula with a maximal domain of validity in the set of Laurent polynomials, i.e., the formula is exact for all trigonometric polynomials in <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s13066019.png" />, a space of dimension <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/s/s130/s130660/s13066020.png" />, which is the maximal dimension possible with a quadrature formula of this form.
+
One cannot take the zeros of the Szegö polynomials $\phi_n$ as nodes (as in Gaussian formulas), because these are all in the open unit disc $\mathbf D$ (cf. also [[Szegö polynomial|Szegö polynomial]]). Therefore, the para-orthogonal polynomials are introduced as $Q _ { n } ( z , \tau ) = \phi _ { n } ( z ) + \tau \phi _ { n } ^ { * } ( z )$, where $\tau \in \mathbf{T}$ and $\phi _ { n } ^ { * } ( z ) = z ^ {n } \overline { \phi _ { n } ( 1 / \overline{z} ) }$. These are orthogonal to $\{ z , \ldots , z ^ { n - 1 } \}$ and have $n$ simple zeros, which are on $\bf T$. The Szegö quadrature formula then takes as nodes the zeros $\xi _ { n , k }$, $k = 1 , \dots , n$, of $Q _ { n } ( z , \tau )$, and as weights the [[Christoffel numbers|Christoffel numbers]]
 +
 
 +
\begin{equation*} \lambda _ { n k } = \frac { 1 } { \sum _ { j = 0 } ^ { n - 1 } | \phi _ { j } ( \xi _ { n k } ) | ^ { 2 } } &gt; 0. \end{equation*}
 +
 
 +
The result is a quadrature formula with a maximal domain of validity in the set of Laurent polynomials, i.e., the formula is exact for all trigonometric polynomials in $\operatorname { span } \{ z ^ { - n - 1 } , \dots , z ^ { - 1 } , 1 , z , \dots , z ^ { n - 1 } \}$, a space of dimension $2 n - 1$, which is the maximal dimension possible with a quadrature formula of this form.
  
 
The Szegö quadrature formulas were introduced in [[#References|[a2]]]. The underlying ideas have been generalized from polynomials to rational functions. See [[#References|[a1]]].
 
The Szegö quadrature formulas were introduced in [[#References|[a2]]]. The underlying ideas have been generalized from polynomials to rational functions. See [[#References|[a1]]].
  
 
====References====
 
====References====
<table><TR><TD valign="top">[a1]</TD> <TD valign="top">  A. Bultheel,  P. González-Vera,  E. Hendriksen,  O. Njåstad,  "Quadrature and orthogonal rational functions"  ''J. Comput. Appl. Math.'' , '''127'''  (2001)  pp. 67–91  (Invited paper)</TD></TR><TR><TD valign="top">[a2]</TD> <TD valign="top">  W.B. Jones,  O. Njåstad,  W.J. Thron,  "Moment theory, orthogonal polynomials, quadrature and continued fractions associated with the unit circle"  ''Bull. London Math. Soc.'' , '''21'''  (1989)  pp. 113–152</TD></TR></table>
+
<table><tr><td valign="top">[a1]</td> <td valign="top">  A. Bultheel,  P. González-Vera,  E. Hendriksen,  O. Njåstad,  "Quadrature and orthogonal rational functions"  ''J. Comput. Appl. Math.'' , '''127'''  (2001)  pp. 67–91  (Invited paper)</td></tr><tr><td valign="top">[a2]</td> <td valign="top">  W.B. Jones,  O. Njåstad,  W.J. Thron,  "Moment theory, orthogonal polynomials, quadrature and continued fractions associated with the unit circle"  ''Bull. London Math. Soc.'' , '''21'''  (1989)  pp. 113–152</td></tr></table>

Latest revision as of 16:59, 1 July 2020

Szegö quadrature formulas are the analogues on the unit circle $\bf T$ in the complex plane of the Gauss quadrature formulas on an interval (cf. also Gauss quadrature formula). They approximate the integral

\begin{equation*} I _ { \mu } ( f ) = \int _ { T } f ( t ) d \mu ( t ), \end{equation*}

where $\mathbf{T} = \{ z \in \mathbf{C} : | z | = 1 \}$ and $\mu$ is a positive measure on $\bf T$, by a quadrature formula of the form

\begin{equation*} I _ { n } ( f ) = \sum _ { k = 1 } ^ { n } \lambda _ { n k } f ( \xi _ { n k } ). \end{equation*}

One cannot take the zeros of the Szegö polynomials $\phi_n$ as nodes (as in Gaussian formulas), because these are all in the open unit disc $\mathbf D$ (cf. also Szegö polynomial). Therefore, the para-orthogonal polynomials are introduced as $Q _ { n } ( z , \tau ) = \phi _ { n } ( z ) + \tau \phi _ { n } ^ { * } ( z )$, where $\tau \in \mathbf{T}$ and $\phi _ { n } ^ { * } ( z ) = z ^ {n } \overline { \phi _ { n } ( 1 / \overline{z} ) }$. These are orthogonal to $\{ z , \ldots , z ^ { n - 1 } \}$ and have $n$ simple zeros, which are on $\bf T$. The Szegö quadrature formula then takes as nodes the zeros $\xi _ { n , k }$, $k = 1 , \dots , n$, of $Q _ { n } ( z , \tau )$, and as weights the Christoffel numbers

\begin{equation*} \lambda _ { n k } = \frac { 1 } { \sum _ { j = 0 } ^ { n - 1 } | \phi _ { j } ( \xi _ { n k } ) | ^ { 2 } } > 0. \end{equation*}

The result is a quadrature formula with a maximal domain of validity in the set of Laurent polynomials, i.e., the formula is exact for all trigonometric polynomials in $\operatorname { span } \{ z ^ { - n - 1 } , \dots , z ^ { - 1 } , 1 , z , \dots , z ^ { n - 1 } \}$, a space of dimension $2 n - 1$, which is the maximal dimension possible with a quadrature formula of this form.

The Szegö quadrature formulas were introduced in [a2]. The underlying ideas have been generalized from polynomials to rational functions. See [a1].

References

[a1] A. Bultheel, P. González-Vera, E. Hendriksen, O. Njåstad, "Quadrature and orthogonal rational functions" J. Comput. Appl. Math. , 127 (2001) pp. 67–91 (Invited paper)
[a2] W.B. Jones, O. Njåstad, W.J. Thron, "Moment theory, orthogonal polynomials, quadrature and continued fractions associated with the unit circle" Bull. London Math. Soc. , 21 (1989) pp. 113–152
How to Cite This Entry:
Szegö quadrature. Encyclopedia of Mathematics. URL: http://encyclopediaofmath.org/index.php?title=Szeg%C3%B6_quadrature&oldid=23067
This article was adapted from an original article by A. Bultheel (originator), which appeared in Encyclopedia of Mathematics - ISBN 1402006098. See original article